[Douglas W. Zochodne] Neurobiology of Peripheral N

This page intentionally left blank

Neurobiology of Peripheral Nerve Regeneration

Peripheral nerve disorders are among the most common neurological problems

that clinicians face, yet few therapies and interventions are available to arrest

or reverse the damage associated with them. Summarizing this important,

but neglected, area of neuroscience, Doug Zochodne addresses the peripheral,

not central, nervous system and its unique neurobiology. He summarizes

current basic ideas about the molecular mechanisms involved in both nerve

degeneration and regeneration and what approaches can be used to address

it experimentally. Heavily illustrated throughout, and including a 32-page

color plate section, this book will serve as a valuable reference for academic

researchers and graduate students.

Dr. Douglas Zochodne is a Professor with Tenure and a Consultant Neurologist

in the Department of Clinical Neurosciences, Hotchkiss Brain Institute at the

University of Calgary. He has recently served (1999–2007) as Editor-in-Chief of

the Canadian Journal of Neurological Sciences. Dr. Zochodne has run an externally

funded research laboratory investigating problems of peripheral nerves since

1989. The work has been funded by the Canadian Institutes of Health Research,

Canadian Diabetes Association, Alberta Heritage Foundation for Medical

Research, and the Muscular Dystrophy Association of Canada.

Neurobiology ofPeripheral NerveRegeneration

Douglas W. ZochodneUniversity of Calgary, Canada

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

First published in print format

ISBN-13 978-0-521-86717-7

ISBN-13 978-0-511-43713-7

© D. W. Zochodne 2008

2008

Information on this title: www.cambridge.org/9780521867177

This publication is in copyright. Subject to statutory exception and to the

provision of relevant collective licensing agreements, no reproduction of any part

may take place without the written permission of Cambridge University Press.

Cambridge University Press has no responsibility for the persistence or accuracy

of urls for external or third-party internet websites referred to in this publication,

and does not guarantee that any content on such websites is, or will remain,

accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

eBook (EBL)

hardback

Dedicated to my wife Barbara and my children Julia and William

Contents

Acknowledgments viii

1 Introduction 1

2 The intact peripheral nerve tree 8

3 Injuries to peripheral nerves 39

4 Addressing nerve regeneration 58

5 Early regenerative events 85

6 Consolidation and maturation of regeneration 133

7 Regeneration and the vasa nervorum 153

8 Delayed reinnervation 170

9 Trophic factors and peripheral nerves 182

10 The nerve microenvironment 206

References 220

Index 267

Color plate section pp. 152–153

vii

Acknowledgments

I owe particular thanks to Barbara Zochodne, my wife, who provided

the encouragement and detailed editing that made this book possible. Barbara,

my daughter Julia and my son William put up with my long hours devoted to

this book.

I am grateful to many members who have worked with me in our laboratory

over the years and who have captured a number of the images that are illus-

trated in this book. In particular I acknowledge Drs. Chu Cheng, Christine

Webber, and David McDonald for their help. I enjoyed the opportunity to work

with Scott Rogers who provided many artistic illustrations for this text. Dr. James

Kennedy, Dr. Cory Toth, Dr. Hong Sun, Dr. X.-Q. Li, Dr. Q.-G. Xu, Dr. Y. Q. Xu,

Lam Ho, Dr. Dan Levy, Dr. Ahmet Hoke, Dr. Valentine Brussee, Wei-Qiao Liu,

G. F. Guo, J. A. Martinez, K. Vanneste, and Noor Ramji have all contributed to this

work among many others I have had the privilege to work with. Brenda Boake has

provided expert secretarial assistance and unswerving support over the same time.

The Peripheral Nerve Society (PNS) has been an intellectual home for many of

us interested in the biology of peripheral nerves and a potent stimulus for me to

proceed with this work.

Our laboratory is grateful for support from the Alberta Heritage Foundation

for Medical Research, Canadian Institutes of Health Research, Canadian Diabetes

Association and Muscular Dystrophy Association of Canada.

“Nature uses only the longest threads to weave her patterns, so each

small piece of her fabric reveals the organization of the entire tapestry.”

(Richard Feynman, The Character of Physical Law, 1965)

viii

1

Introduction

This book is about peripheral nerves, their unique biology and how they

repair themselves during regeneration. The biology of the peripheral nervous

system is not often considered on its own. Much has been learned about the

neurosciences of peripheral nerves, specifically during injury and regeneration,

but it is my sense that some of this new and exciting information should be

consolidated and considered in an overview.

Without nerves, specifically peripheral nerves, there is no movement, no

sensation. Peripheral nerves are the essential connections between the body,

brain, and spinal cord. The “peripheral nervous system (PNS)” distinguishes itself

from the “central nervous system (CNS)” on many levels. Peripheral axons reside

in many types of local environments including muscles, connective tissue, skin,

and virtually every organ of the body. This reach extends into the meninges that

surround the brain, a surprising fact to some. Moreover, peripheral neurons are

very different from their CNS counterparts in how they respond to injury or

disease, in which cells they partner with and in what axon trees they support.

For example, a sensory neuron in the lumbar dorsal root sensory ganglion is

required to maintain and support distal axon branches that can extend a meter

or more to the skin of the toe. Only a small proportion of CNS neurons have

comparable outreach and demands placed upon them.

“Neuropathies,” of which there are a large number, are simply disorders of

peripheral nerves. A neuropathy might be focal (also known as a mononeuro-

pathy) and involve only a single peripheral nerve, or it might involve peripheral

nerves widely (polyneuropathy). Despite being very common problems, compar-

able in prevalence with stroke and Alzheimer’s disease, they are not widely

understood by patients, health care providers or neuroscientists! Polyneuropathy

can be detected in approximately half of all diabetic subjects, an important issue

1

to consider in this day of dramatically rising Type 2 diabetes prevalence.

Diabetic neuropathy itself, without considering all other forms of neuropathy,

has a prevalence of over ten times that of MS.

Consider a few important points. A patient with severe peripheral nerve

disease, such as Guillain–Barre syndrome (GBS) can, during the acute phase be

completely “locked in,” or unable to move a single limb muscle or eye muscle.

This patient may require a ventilator to breathe. He may also have lost any

sensation to light touch, pain, or temperature. Despite these severe deficits,

however, cognition may well be fully retained because the disorder does not

involve the brain. It is difficult to conceive of being “locked in” while being fully

conscious unless one has suffered from GBS or a comparable disorder of the

peripheral nervous system. The reader is referred to books written by patients

who have suffered and recovered from GBS [373,374]. GBS is an autoimmune

inflammatory polyneuropathy, sometimes triggered by infections or vaccinations

that took place 2–3 weeks earlier. Different types of GBS are recognized. Yet it is

the unique neurobiology of the nerves damaged during GBS that will most impact

how a patient might fare. The most common form of GBS, also known as the

classical demyelinating type, involves only the myelin sheath of the peripheral

nerve. The underlying axon tree remains intact despite having been rendered

nonfunctional by the loss of its myelin sheath. Remyelination of the axons is

expected and can be associatedwith a complete recovery of paralysis and sensation.

Alternatively, a type of GBS recently termed AMSAN (acute motor and sensory

axonal neuropathy) primarily attacks the axons and spares the myelin. Recovery

is dictated by the rate and likelihood (certainly not guaranteed) that axons will

regrow from the injury site to their correct original target, e.g., a small foot

muscle, a touch receptor in the finger. The unfortunate result is very limited,

delayed, or absent recovery in this severe form of GBS. While axons might be

expected to regrow at the rate of about an inch per month in order to reach their

targets, this likelihood falls dramatically with time. These limitations will be

discussed in subsequent chapters. The tragedy is that, in some instances, severe

GBS that primarily attacks axons may not recover at all.

Consider the story of “Nancy B.,” a young woman who garnered national

attention in Canada because of her peripheral nerve disorder. She developed severe

axonal GBS that rendered her “locked in” and ventilator bound without any

improvement over 2½ years. While being perfectly lucid about her condition, she

made the decision to have her health care workers withdraw her ventilator

support. Without this support, she did not survive. The story generated wide-

spread discussion about ethical issues surrounding the maintenance of life sup-

port in patients who do not have a chance for recovery. Figure 1.1 is an image of a

patient who required intensive care unit hospitalization for 1 year because of GBS.

2 Introduction

Fortunately, many neuropathies do not render disability as severe as that

experienced by “Nancy B.” They do, however, impose their own range of disabil-

ities and interference with quality of life. Some are associated with significant

loss of function. Consider the diabetic patient seen by the author who developed

a focal mononeuropathy of his ulnar nerve at the elbow. This lesion rendered

significant, though not complete, hand wasting and weakness. The patient,

however, had been a professional tennis player and could not accept his inability

to play. The neuropathy compounded underlying depression in this patient and

led to suicide. Figure 1.2 is an image of a patient with a focal sciatic nerve injury

lesion from a buttock firearm wound, rendering paralysis of muscles in the leg

and loss of sensation in the foot.

Loss of sensation is associated with loss of the ability to detect skin and soft

tissue injuries. Patientsmay develop skin ulcers fromunrecognized injury to their

feet (e.g., stepping on a nail or damaging their skin from overly tight footwear).

In some cases these injuries are associated with additional damage, infection,

Figure 1.1 A patient with severe Guillain–Barre syndrome with axonal damage and

paralysis of all of his limbs. He required intensive care unit support to breathe for

a period of 1 year before recovering.

3Introduction

and the need for amputation. Polyneuropathy is a leading cause of lower limb

amputation in diabetes. Loss of sensation to position (proprioception) also con-

tributes to falls and injury because it is impossible for a patient to tell where the

limbs are in space. Finally, peripheral nerve damage of all types is frequently

associated with a severe and debilitating type of pain known as “neuropathic

pain.” Neuropathic pain can render patients unable to walk, work, sleep, or

enjoy life. While this text does not directly address neuropathic pain, full and

effective regeneration of the peripheral nervous system usually extinguishes it.

Unlike many other disorders, neuropathies impose a burden of neurological

deficit that requires nerve regeneration, irrespective of what caused the damage.

Therapy for an active peripheral nerve disease, or microsurgical repair of a

transected peripheral nerve trunk (we use the term “trunk” to refer to a periph-

eral nerve branch containing hundreds to thousands of individual axons and

their supporting cells) may address the inciting lesion that caused damage, yet

it is regeneration that must ensue to restore proper function. For example,

vasculitic neuropathy is a disorder that damages peripheral nerve axons through

inflammation of nutrient feeding vessels of the peripheral nerve trunk. It may be

“cured” with a course of immunosuppressive therapy, a treatment that arrests the

inflammation but does not restore function. This is unsatisfactory tomany patients

who suffer from neuropathy; previously damaged axons must now regrow to

Figure 1.2 A patient with a right buttock bullet wound that damaged his underlying

sciatic nerve trunk. The severe axonal disruption caused by the injury resulted in

permanent paralysis and atrophy of muscles in the thigh and below the knee with

sensory loss in the foot and leg. (From Tinel [681].)

4 Introduction

reverse the deficits that have developed. At the time of this writing, specific therapy

designed to coax more complete and effective recovery of nerves is unavailable.

In the neurosciences literature, peripheral neurons have been highlighted

as examples of neurons whose axons can regenerate, unlike those of the CNS.

Indeed, significant excitement has resulted from findings that injured CNS

neurons in the spinal cord can regenerate into and through peripheral nerve

grafts. Peripheral neurons have been seeded on substrates of CNS myelin to

demonstrate its property to inhibit regrowth. Without diminishing the import-

ance of these findings, however, they do not address the realities of peripheral

nerve disease. In neuropathies, the fate of axons regenerating in their own

peripheral microenvironment is the important consideration. Recovery is slow,

and if the distances to the target tissues are long, regeneration may never occur.

Such catastrophic failure occurs despite the fact that axons have a “denervated”

distal stump into which to grow. Distal to an axonal injury, axons undergo the

process of “Wallerian degeneration” in which disconnected branches are phago-

cytosed and eventually disappear. They leave behind a denervated distal stump

that includes a connective scaffold and supporting Schwann cells. In this text,

my intent is to dispel the idea that peripheral neurons serve simply as a model

for understanding issues in CNS regeneration. Rather, I seek to convince the

reader that PNS disease poses its own unique burdens on a substrate of neurons

and supporting cells and has its own, separate but compelling regeneration

issues. It deserves an equal and focused place at the neurosciences research table.

Exciting new aspects of peripheral nerve behavior challenge traditional

concepts. One such example involves how axons interact with their basement

membranes. Extracellular basement membrane constituents of nerve trunks

expose specific ligand (e.g., the RGD, or Arg–Gly–Asp tripeptide sequences)

moieties that interact with integrin receptors of adjacent axons. Local signaling

cascades within axons are triggered by this interaction. These cascades have the

capability of altering growth cone behavior and influencing regeneration yet

may be completely independent of changes within their cell body. The idea that

growth cones and axons locally might signal and react is novel. Not only do such

signals influence axon behavior but in this scenario they also alter local protein

synthesis, previously considered the sole purview of the cell body. To coordinate

regeneration, the axon and perikarya also sense that there is injury, alter the

pattern of nuclear gene expression in the cell body and change the repertoire of

proteins they transport down the axon to the injury site. They do this while

managing to signal local axons to synthesize regeneration-related molecules.

How the whole family of regeneration molecules is coordinated between local

synthesis or transport from the cell body is not known at this time. For example,

local axon synthesis might act as a “rapid response” program for injured axons,

5Introduction

later supplemented by reinforcements shipped from the cell body. One might

imagine that nerve surgeons could one day be capable of implanting regener-

ation conduits with graded release of signals that would “shore up” such local

axonal events.

There are intriguing discoveries in how surrounding and supporting cells of

the peripheral nervous system interact with damaged neurons. For example,

a remote injury of a sensory neuron axon branch (an “axotomy” is a transection

of an axon branch of a neuron) is sufficient to send information to its parent

cell body in the dorsal root ganglion (DRG) up to a meter away. In response,

perineuronal satellite cells in the ganglion that have not been directly involved

in the injury dramatically change their phenotype. Satellite cells are cousins of

Schwann cells and both cell types appear in sensory ganglia. Satellite cells,

however, are interesting cytoplasmic poor cells that closely surround individual

sensory neurons. They exhibit a dynamic form of life and death plasticity with

ongoing apoptosis and division within “stable” ganglia. Their plasticity contrasts

sharply with the apparent immutability of their neighbor neurons. Within a

defined time course, satellite cells enlarge and proliferate around closely associ-

ated but axotomized neurons. How neurons communicate with these important

and pervasive neighbors is unknown. Communication between neurons and

satellite cells is likely to be reciprocal. For example, satellite cells are known to

provide trophic molecules such as CNTF (ciliary neurotrophic factor) to support

neurons and protect them from injury.

In contrast to sensory neurons, injured motor neurons have cell bodies

that are official residents of the CNS, in the anterior horn of the spinal cord.

When their remote axons are injured, their dendrites in the gray matter of the

spinal cord retract. This accompanies “synaptic stripping” of their connections

with other spinal cord neurons. How the loss of these dendritic connections

occurs and how they might be restored is uncertain.

The Schwann cell (SC) is a type of glial cell, unique to the peripheral nervous

system, that supports all types of axons: sensory, motor, and autonomic. Its roles

can be surprisingly multifaceted and differ from those of their CNS myelinating

counterpart, the oligodendrocyte. Early after nerve injury, SCs can serve as local

inflammatory cells by generating cytokines, inducible nitric oxide synthase

(iNOS; an inflammatory enzyme that generates nitric oxide (NO)) and other

inflammatory molecules well before macrophages from the bloodstream enter

the nerve and assume this function. SCs can interconvert from “stable” myeli-

nating phenotypes to highly plastic proliferating and migrating cells that may

direct appropriate and directional axon regrowth after injury. They offer the

peripheral nerve a range of trophic molecules but not necessarily simultan-

eously. In other words, SCs appear to have a sense of timing and coordination

6 Introduction

in what they synthesize. While the exact mechanism is unclear, there is evidence

that they accurately guage their local microenvironment and respond accord-

ingly. Moreover, like the neuron and its perineuronal satellite cell in the

ganglion, there is intimate and bi-directional talk between the axon and the

SC. Cross talk is likely critical during regeneration but may also be a feature of

normal uninjured nerve trunks.

Axons elaborate neuregulins, potent molecules capable of altering SC protein

synthesis, myelin synthesis, and their likelihood to proliferate. Neuregulins, in

turn, instruct SCs to synthesize a series of molecules, such as neurotrophins that

encourage axon regrowth in a highly directional manner. When peripheral

nerve trunks are transected, the tension normally present within them causes

the distal and proximal stumps to retract from one another. While such lesions

are incompatible with full axon regrowth, the stumps can reconnect by a

connective tissue bridge if they are apposed to one another in a graft or conduit.

From the proximal stump of the transected nerve, axons then enter the bridge

and begin to grow across it. These early events offer fascinating opportunities

as to how axons navigate new territories. One interesting finding is that SCs

appear to lead axons through complex, three-dimensional trajectories. Their

relationship, closely linked with local trails of laminin, is so encompassing and

intimate that it might be called the “axon-SC” dance! SC partnership is critical to

the success of axon regrowth. Not surprisingly, a group of colleagues interested

in SCs call themselves the “Friends of Schwann”!

The purpose of this text is to emphasize the unique structure, plasticity, and

challenges of regrowing peripheral nerves. Excepting focal neuropathy associ-

ated with direct injury to the nerve, neuropathies are not addressed directly.

We refer readers to other comprehensive texts addressing peripheral nerve

disorders and peripheral nerve surgery [159,421].

We begin by examining properties of the peripheral nervous system in nerve

trunks that house axons, Schwann cells and other tissue components, and

ganglia that house cell bodies, or perikarya. Next we address how peripheral

nerves are injured by trauma. What are the resulting injuries, their implications,

and the barriers to regrowth? We then address experimental approaches to

peripheral nerve regeneration. We ask how does nerve regeneration evolve

through its early events and later consolidation? Special consideration will

be given to the microvascular supply and its impact on regenerative events.

Finally, we address important aspects of regrowth: the impact of long-term

denervation, the actions of growth factors and molecular barriers of regrowth.

It is my intent that this text might be a project in evolution, consolidating

what has been discovered to date, as well as being a catalyst for new ideas and

approaches toward resolving the burden of peripheral nerve damage.

7Introduction

2

The intact peripheral nerve tree

A thorough appreciation of the unique anatomy of the peripheral nervous

system is essential in understanding how it regenerates. This information,

already described in several texts, is nonetheless summarized here to prepare

the reader for later chapters. There are many facets to peripheral nervous system

anatomy that have a bearing on its response to injury including the multiplicity

of neuron subtypes and the qualities of their housing.

Overall structure

The peripheral nervous system (PNS) is complex. The peripheral nerve

“trunk” refers to a cable of tissue in which hundreds to thousands of axons may

travel. Peripheral nerve trunks form connections from the brain and spinal cord

to all skeletal muscles in the body through motor axons. They also connect all

sensory organs to the brain and spinal cord through sensory axons (Figure 2.1).

Finally, they connect the CNS to smooth muscles, sweat glands, blood vessels,

and other structures through axons of the autonomic nervous system. Axons

traveling through nerve trunks originate from cell bodies of neurons (perikarya)

in the brainstem, spinal cord, and ganglia. Motor neuron cell bodies found in

cranial motor nuclei supply the head and neck, and those in the anterior gray

matter horn of the spinal cord supply the limbs and trunks. Motor neuron cell

bodies have their greatest numbers in the cervical and lumbar enlargements of

the spinal cord so that they can supply the large number of muscles in the upper

limbs and lower limbs, respectively. Sensory neuron perikarya are found in

cranial sensory ganglia and paraspinal ganglia from approximately T1 through

to L1 levels. Autonomic neurons are housed in a variety of sites: cranial and

cervical ganglia, paraspinal sympathetic ganglia, and a variety of ganglia in the

8

Figure 2.1 An illustration of the peripheral sensory nerve territories of the human body.

Left panel, anterior body: A – greater auricular nerve; B – anterior cutaneous nerve of neck;

C – supraclavicular nerves; D – medial cutaneous nerve or arm and intercostobrachial

nerve; E – medial cutaneous nerve of the forearm; F – radial nerve; G – median nerve;

H – ulnar nerve; I – iliohypogastric nerve; J – genital branch of genitofemoral nerve;

K – scrotal branch of perineal nerve; L – oburator nerve; M – lateral cutaneous nerve of

calf; N – superficial peroneal nerve; O – sural nerve; P –medial and lateral plantar nerves;

Q – deep peroneal nerve; R – sapenous nerve; S – intermediate and medial cutaneous

nerves of the thigh; T – lateral cutaneous nerve of the thigh; U – dorsal nerve of penis;

V – femoral branch of genitofemoral nerve; W – ilioinguinal nerve; X – lateral

cutaneous nerve of forearm; Y – lateral cutaneous nerve of arm; Z – axilllary nerve.

Right panel, posterior body: A – greater and lesser occipital nerves; B – anterior cutaneous

nerve of neck; C – axillary nerve;D –medial cutaneousnerve of armand intercostobrachial

nerve; E – lateral cutaneous nerve of forearm; F – medial cutaneous nerve of forearm;

G – posterior cutaneous nerve of forearm; H – radial nerve; I – median nerves; J – ulnar

nerve; K – inferior medial clunical nerve; L – obturator nerve; M –medial cutaneous nerve

of thigh; N – lateral cutaneous nerve of calf; O – sural nerve; P – calcaneal branches of sural

and tibial nerves; Q – superficial peroneal nerve; R – saphenous nerve; S – posterior

cutaneous nerve of thigh; T – lateral cutaneous nerve of thigh; U – inferior lateral clunical

nerve; V – iliohypogastric nerve; W – lower lateral cutaneous nerve of arm; X – posterior

cutaneousnerveof arm; Y– supraclavicularnerves; Z –greater auricularnerve. (Illustration

by Scott Rogers, based on previous illustrations by Haymaker and Woodall [254].)

9Overall structure

abdomen that include the celiac, mesenteric, para-aortic, hypogastric, and others.

The enteric nervous system includes a large number of neurons within the walls

of the gastrointestinal system.

Nerve “trunks” originate through a confluence of branches that supply them.

Motor neurons send axons to nerve trunks through the ventral roots of the

spinal cord and motor branches of cranial nerves. Sensory neurons have an

initial single branch that emerges from the cell body and then divides into two

branches, an arrangement called “pseudounipolar” (see below). From the initial

single branch, a central branch is directed to the spinal cord entering the dorsal

horn and posterior columns and a peripheral branch is sent to the nerve trunks.

The peripheral sensory branch in the dorsal root joins motor axons from the

ventral roots as they exit the neural foramina of the bony spinal column and

together form the mixed spinal nerve. It is at this site that the meningeal sheath

(dura and arachnoid) surrounding the spinal cord and its roots forms pockets or

sleeves that blend into the epineurial sheath of the peripheral nerve. Mixed

spinal nerves then send branches posteriorly to innervate the paraspinal muscles

and anteriorly where they form the major nerve trunks of the body. From both

the cervical enlargement and the lumbar enlargement, anterior spinal nerve

branches merge and intermingle to form the brachial and lumbosacral plexus,

respectively. From each plexus, peripheral nerves are then formed from a mix-

ture of motor, sensory, and autonomic axons arising at different root levels.

Despite the mixture, most major nerve trunks include axons from only a few

spinal root levels. Thus, for example, the human median nerve is composed

almost exclusively of axons from the C8 and T1 spinal root level, while the

musculocutaneous nerve arises from C5 and C6.

Given this overall arrangement then, most major nerve trunks house a variety

of axon types. For example, there are likely no “pure” motor nerve trunks since

nerves associated with muscles include both motor axons and a large comple-

ment of sensory axons sensitive to muscle pain, or to stretch in muscle spindles.

Cutaneous nerves do not include motor axons but do contain both autonomic

and sensory axons. In any major peripheral nerve trunk, therefore, there are

larger myelinated axons (a motor axons and large Aa sensory axons), small

myelinated axons (smaller Ab and Ad sensory axons, g motor axons – see below),

and unmyelinated axons (C sensory and autonomic axons). Classical histological

approaches do not distinguish whether myelinated axons are motor or sensory

or whether unmyelinated axons are sensory or autonomic. Thus, when a trans-

verse section of a peripheral nerve trunk is examined, it is not possible to

identify what class a given axon may belong to. In humans the sural nerve is

most often harvested for diagnostic purposes and its axons are sensory or

autonomic only.

10 The intact peripheral nerve tree

Immunohistochemical labeling does allow axons to be more accurately

distinguished. Labeling can be applied for neurotransmitter enzymes or peptides

specific for neuron subtypes, e.g., choline acetyltransferase (ChAT) for motor

axons, tyrosine hydroxylase (Th) for sympathetic autonomic axons. Substance

P (SP), is specific for unmyelinated sensory nociceptive (pain transmitting) axons.

Other unmyelinated sensory axons are labeled by a lectin tag known as IB-4 and

express receptors sensitive to the growth factor GDNF (glial derived neurotrophic

factor). They have also been labeled NGF unresponsive or Trk A negative small

afferents.

Large nerve trunks frequently course as “neurovascular bundles” together

with blood vessels. Bundles can be identified in humans at the inguinal liga-

ment, popliteal fossa, thoracic inlet, and other areas. This kind of partnership,

however, is not invariable and nerve trunks can find their way separately to

the muscles and sensory organs they innervate. At periodic intervals along the

nerve trunk, they are supplied by feeding branches from adjacent blood vessels

forming the vasa nervorum, or the blood supply of the peripheral nerve. Vasa

nervorum form into a plexus on the epineurial surface of the nerve that, in turn,

regularly penetrates to supply deeper structures. Longitudinal vessels arising

from the epineurial plexus also enter the nerve trunk and travel within it.

Thus, between penetrating plexus arterioles and longitudinal supplying vessels,

portions of peripheral nerves not directly attached to obvious feeding vessels

are nonetheless well perfused. For larger peripheral nerve trunks, such as the

sciatic nerve in the midportion of the thigh, some portions may be especially

vulnerable to interruptions in this supply, or ischemia. These are so-called

“watershed” zones of the nerve trunk where, in the absence of adjacent feeding

vessels, they instead rely on remote supply from an intact epineurial plexus or

a longitudinal vessel.

Structure and characteristics of the nerve trunk

The structure of the peripheral nerve trunk is unique and dramatically

differs from that of the brain, the spinal cord, or the optic nerve. Its tensile

strength, important for the routine bending and twisting encountered during

normal limb movements, is accounted for by its collagen sheath structure.

Collagen polymers are oriented longitudinally parallel with the trunk cable

and provide resistance to stretch, and protection from compression. When

tissue pressure rises within the peripheral nerve trunk, its structure also offers

compliance. Compliance means that rises in local pressure can occur with-

out occluding the local vascular supply and that some stretching can occur

without severing axons.

11Structure and characteristics of the nerve trunk

The epineurium or epineurial sheath is the outer layer of the peripheral nerve

trunk and includes the collagen tissue sheath, a plexus of blood vessels, lymph-

atic vessels, some resident macrophages, fibroblasts, and mast cells. Its thickness

can be quite variable and it sometimes includes adipose tissue (Figures 2.2, 2.3).

The epineurial sheath has little connection to adjacent tissues allowing nerves to

normally slide and move during limb movement [43]. It consists predominantly

of Type I collagen but elastic fibers are also present, and they are largely

longitudinally oriented. Wavy patterns of the inner epineurial collagen and

the axons allow the nerve to be extended or stretched to some degree [43].

The epineurial vasa nervorum of the peripheral nerve trunk lack a specific

barrier to blood borne molecules. Thus, for example, classical experiments using

intravascular injections of Evans blue albumin noted that it fully permeates out

through vessels into the tissues of the epineurial sheath, a permissive state that

may be relevant during inflammatory disorders of nerves.

Local blood flow in the epineurium is high, at least two to three times that of

the interior, or endoneurium (see below) of the nerve trunk. When peripheral

nerves are examined at surgery, the dense plexus of the vasa nervorum can

thereby be observed on the surface of the nerve trunk with highly irregular

and tortuous contours. Some of this irregularity or redundancy also gives blood

vessels flexibility in dealing with stretch or twisting of the nerve trunk without

compromising their ability to supply the nerve. It is interesting, for example, that

peripheral nerve surgeons can “mobilize” or free significant lengths of peripheral

perineurium

epineurium

endoneurium

Figure 2.2 An illustration of the overall organization of the peripheral nerve

trunk and its compartments. (Illustration by Scott Rogers.) See color plate section.


nerves from their surrounding tissue without damage. A rich anastomotic supply

of blood vessels allows such leeway. Arteriovenous (AV) shunts, direct connec-

tions between arterioles and veins, occur in the epineurial plexus.

The epineurial blood supply of the peripheral nerve trunk also has unique

physiological qualitities that are discussed in more detail later in Chapter 7. One

property is its lack of autoregulation, distinguishing it from cerebral blood flow.

Vasa nervorum also possess vasoactive properties. They are richly innervated

by both sympathetic adrenergic terminals and peptidergic endings that can

influence downstream flow of the blood vessels they innervate. For adrenergic

terminals, activation results in vasoconstriction and declines in downstream

blood flow. Specific segments of epineurial arterioles probably control flow, for

example, at vessel junctions as “precapillary sphincters.” Thus arteriolar segments

with adrenergic terminals might be critical sites of “control” or “gateways” for

downstream blood supply to endoneurial structures. Peptidergic nerve endings

on vasa nervorum, containing calcitonin gene-related peptide (CGRP), SP, and

others, arise from sensory axons. These have the capacity to dilate epineurial

vessels and this may be important during nerve inflammation. Both adrenergic

terminals and peptidergic terminals on vasa nervorum provide ongoing “tone”

from a basal level of activity. In an otherwise intact epineurial vascular bed then,

Figure 2.3 A photomicrograph of a sciatic peripheral nerve trunk. Note the presence

of two major fascicles with myelinated axons within the endoneurium. The section is

a low-power view of a semithin section that was harvested from an adult rat, fixed,

embedded in epon and stained with toluidine blue.


interruption of “normal” adrenergic tone results in vasodilatation, whereas

blocking peptidergic actions (mainly CGRP as it is the more potent vasodilator)

causes vasoconstriction.

Overall then, while epineurial arterioles are controlled by nerve terminals,

whether this innervation extends to venules or AV shunts is uncertain. The

epineurium is self-innervated. By this, we mean that small unmyelinated axons

from the endoneurium of the parent nerve trunk travel outward into the

connective tissue of the epineurium. Some, but not all of these axons innervate

blood vessels, as discussed above. Termed “nervi nervorum” these small axons

have been thought to generate pain in some types of peripheral nerve damage

and inflammation [25,53]. Local mast cells that contain histamine and serotonin

are often identified as residents near epineurial vessels. Lymph vessels are

present in the epineurium, but not in other areas of the nerve [346].

The perineurium is a laminated cylindrical layering of specialized cells that is

found deep to the epineurium and that surrounds and protects the endoneurial

fascicles (Figures 2.2, 2.3). The perineurial cells are interleaved and intercon-

nected by tight junctions (zona occludens) and gap junctions (zona adherens)

forming layers separated by collagen fibrils (Type IV). Perineurial cells also

contain pinocytotic vesicles that probably contribute to regulated forms of

transcellular transport [43]. They have a basal laminae consisting of collagen,

fibronectin, laminin, and glycosaminoglycans. While there has been debate, it is

now agreed that perineurial cells are of fibroblast origin [77].

The tight junctions between perineurial cells and the presence of vesicles

within them form part of the blood nerve barrier (BNB), a barrier to blood-borne

constituents from entry into the endoneurial fascicles. This barrier, like that

of the brain (blood brain barrier), helps to protect the endoneurium from

potentially toxic serum proteins, micro-organisms, and other constituents. Thus,

after the indicator Evans blue albumin is given, tissues within the intact peri-

neurium are unstained because the protein does not penetrate the barrier

[534,535]. The BNB is somewhat leakier than its central counterpart, the blood

brain barrier, and it protects axons, not neurons. It is more than just a barrier

to proteins such as Evans blue albumin but also excludes small molecules and

ions. Weerasuryia, Rapoport, and others have described specific quantitative

approaches to studying the BNB [575,730,733,734]. The perineurium extends

the length of the peripheral nerve and connects with the pia and arachnoid

layers of the brain and spinal cord as nerve roots travel toward the CNS.

Perineurial cells also extend to help form the inner capsule of the dorsal root

ganglia and outer layer of sensory organs such as Pacinian corpuscles [43].

The endoneurium might be considered the most important part of the nerve

trunk since it contains axons and their supporting SCs. It also contains mast


cells, resident macrophages, some fibroblasts, and blood vessels (Figures 2.2, 2.3).

Types I and III collagens are found within the endoneurium. Endoneurial fluid,

in turn, has a unique composition that may be hypertonic based on measures

using energy dispersive spectrometry. In 100 picoliter samples of endoneurial

fluid its electrolyte composition was calculated as: Naþ 179, Cl– 131, Kþ 21mEq/L

[490]. In most peripheral nerve trunks, there may be a variable number of

distinct endoneurial compartments, also known as fascicles. Each is surrounded

by its own perineurial sheath and the whole nerve trunk is then termed multi-

fascicular. For example, the median nerve at the wrist in humans may contain

upwards of ten such individual fascicles. What determines the number of

fascicles in a nerve trunk is unknown. During regeneration of nerves “mini-

fascicles” can be formed and they often precede the consolidation and formation

of more classical mature nerve trunk structure (see Chapter 6).

Endoneurial blood vessels are largely capillary, although some arterioles

are found coursing through the fascicle (Figure 2.4). Capillary endothelial cells

may be associated with contractile pericytes. Most endoneurial vessels are not

innervated and the endothelial cells are connected by tight junctions. Thus,

like the perineurial cells, the endothelium of endoneurial vasa nervorum also

contributes to the blood nerve barrier.

Figure 2.4 A photomicrograph showing the structures within the endoneurium

of a nerve trunk fascicle. The arrow points to a myelinated axon. Note that there

are populations of larger and smaller myelinated fibers. The section also shows

endoneurial capillaries (c) and an arteriole (a). The section is a medium-power view

of a semithin section as in Figure 2.3.


Individual axons may move from one fascicle to another as they travel from

proximal to distal. Small “branching” fascicles that contain these mingling

axons are commonly observed in transverse sections of peripheral nerves. Axon

mingling, however, is incomplete and nerve trunks maintain an overall topo-

graphical distribution of axons within them. As a result, axons retain locations

within specific fascicles from proximal to distal and there are specific groupings

destined for individual muscles. The issue is important because knowledge of

this microscopic anatomy of an individual nerve during electrophysiological

studies or surgery can help to predict how repair should be approached.

Two major types of axons are identified within the endoneurium (Figures 2.5,

2.6). Myelinated axons are larger in caliber and surrounded by a lamellar

lipid myelin sheath. Along the myelinated axon are single Schwann cells (SCs),

each forming a length of myelin that form internodes, in turn separated by

nodes of Ranvier (Figure 2.7). In a normal human nerve, internodes range from

300–2000 microns in length, depending on the axon size and a very general guide

is that the internodal length approximates 100� the axon diameter [194]. Nodes

of Ranvier expose a portion of the axon membrane containing sodium channels

that allow transmission of action potentials by “saltatory” transmission.

In a human nerve, such as the sural cutaneous nerve, myelinated axons are

distributed into two main size categories (also known as a bimodal distribution

Figure 2.5 A high-power, oil immersion photomicrograph of endoneurial constituents

including a mast cell, arteriole, and myelinated axons. Note that some axonal

ultrastructural components can be resolved in the large myelinated axons (most likely

microtubules and mitochondria) and in the white box, groups of unmyelinated axons

can just be resolved. This is a semithin section as in Figure 2.3.


of fiber sizes). The size distribution can be appreciated by a histogram where

the number of axons in given size ranges are plotted (numbers or density of

axons on the y-axis and size categories on the x-axis). Larger myelinated axons,

identified by a peak fiber diameter range of 5–12 microns, represent Aa sensory

Figure 2.6 An electron micrograph showing two medium caliber myelinated axons

with some SC cytoplasm around each. In the lower left part of the image, two clusters

of unmyelinated axons (Remak bundle) is seen.

Figure 2.7 A single teased myelinated axon is shown above with the small arrows

pointing to nodes of Ranvier (top image). The segment between the arrows is called an

internode. Below is a higher power view showing a node. Note the darkened axon in

the node region from more closely packed microtubules and neurofilaments. The

lower image is a longitudinal semithin toluidine blue section examined under an oil

immersion lens.


axons that serve light touch, vibration, and joint position. Their conduction

velocities are in the range of 40–80m/s or higher. Ab axons are smaller, inter-

mediate axons with slower conduction velocities. A population of small myeli-

nated axons, also known as Ad, forms a distinct second peak, and these axons

subserve nociception, thermal sensation, and perhaps other functions. Their

fiber diameters range from 1–4microns and their conduction velocities range from

12–30m/s. In adult humans the mean myelinated fiber density is approximately

12000–15000/mm2 in the proximal human sural nerve and 7000–10000/mm2 in

its distal portions [156]. For motor fibers, a motor axons are only slightly smaller

than Aa sensory axons and g motor axons are in the mid range.

Unmyelinated axons represent the second major category of axons in the

endoneurium. Several of these axons are usually associated with a SC in a unit

known as a Remak bundle (Figure 2.8). The number of unmyelinated axons

associated with a single SC in a Remak bundle ranges from 1 to 20 or more.

The mean number is approximately 6 fibers per SC or up to 36 in a rat, with

fewer numbers in man [478]. These SCs differ from those associated with myeli-

nated axons and they form overlapping chains to incorporate axons along their

length. A given axon, however, may not accompany the same neighbor axons

in SC units along its trajectory. Axons may be resorted along their length into

various Remak units. SCs may support different kinds of axons as well. In the

Figure 2.8 An electron micrograph of a Remak bundle of several unmyelinated

axons (arrows) associated with a single SC.


cervical sympathetic trunk of the rat, for instance, the same SC may enwrap

both upwardly projecting preganglionic axons and downwardly projecting

postganglionic axons [802]. The size distribution of unmyelinated axons is uni-

modal, or having one peak, with fiber diameters that range from 0.4–1.2 microns

and conduction velocities from 0.5–2m/s. In humans, the mean unmyelinated

fiber density is approximately 30000/mm2 [522,523].

The endoneurium also includes subperineurial Renaut bodies. These are large

pale structures, highly variable in number and distribution, that consist of

oxytalin fibril connective tissue, components of elastic fibers. Renaut bodies

are thought to arise from perineurial cells.

The transitional zone (TZ) refers to the connection between the peripheral

and central nervous system [43]. It may include protrusions of CNS material into

a PNS root, or insertions of PNS material into the CNS. Astrocytic processes

separate axons as they enter through the glial limitans, the outer border of

the CNS and basement membranes add barrier properties to segregate the

endoneurium from the CNS.

Axon ultrastructure

The intact mature peripheral nerve axon is constructed of a scaffold

of proteins consisting of microfilaments, neurofilament intermediate proteins,

and microtubules all surrounded by an axolemma (Figure 2.9). Microfilaments

are expressed in the subaxolemmal cytoskelton, whereas neurofilaments and

microtubules are distributed throughout the axon. Peripheral neurons express

five major types of intermediate filament proteins: three neurofilament sub-

units, peripherin, and a-internexin (for review see [375]). The axolemma is a

three-layered cell membrane about 8nm thick that is anchored to the subjacent

axoplasmic cortex by molecules such as ankyrin, fodrin, actin, and A-60 [43].

It therefore also houses several organelles that include mitochondria and poly-

ribosomes but does not include Golgi organs or rough endoplasmic reticulum

[43] (van Minnen, personal communication). Mitochondria, identified as flat-

tened tube-like double membrane structures (0.1–0.3 by 0.5–1.0 microns in size),

are transported anterogradely and retrogradely and their density is higher

in smaller caliber axons. The axoplasmic reticulum is a fine meshwork of inter-

connected tubules that connect proximally to rough endplasmic reticulum and

Golgi organs in the perikaryon or cell body. Dense lamellar bodies and multi-

vesicular bodies are thought to represent lysosomes and residual bodies whereas

vesiculotulular profiles likely represent vesicles undergoing rapid anterograde

axoplasmic transport. Membranous cisterns can also be identified in axons and

they include endosomes.

19Axon ultrastructure

Neurofilaments are class IV intermediate filaments. They are stable polymers,

10nM in diameter, that contain three subunit proteins, termed light (68kD,

NfL), medium (145kD, NfM), and heavy (200kD, NfH). In transverse sections,

neurofilaments are spaced regularly and they provide a scaffold or skeleton that

gives the axon caliber or bulk. Each neurofilament subunit consists of a globular

“head” region, an a-helical “rod” region, and a globular C-terminal “tail” exten-

sion. NfM and NfH have sidearm extensions from their tail zones that help to

determine their spacing. Thus, the separation of the polymer from its neighbor

influences the overall size of the nerve. Sidearms from NfH and NfM tail regions

contain KSP repeat sequences that are subject to phosphorylation, a modification

that influences neurofilament spacing and thereby axonal caliber. Moreover,

neurofilament polymers in the cell body are relatively poorly phosphorylated,

whereas axonal neurofilaments are heavily phosphorylated, a process that occurs

after the protein or its subunits are transported from the cell body. At nodes

of Ranvier, alterations in neurofilament phosphorylation likely account for

the closer spacing and axonal narrowing that normally occurs (Figures 2.7, 2.9).

Neurofilaments are also subject to glycosylation [456].

Several animal models have been studied that lack all neurofilaments or one

or more neurofilament subunits. Mice generated by Eyer and Peterson [176]

completely lacked axonal neurofilaments, but survived normally, and had axon

myelination. The model was discovered by serendipity as a result of replacing the

Figure 2.9 An electron micrograph of a myelinated axon illustrating its

ultrastructure. The enlarged box identifies neurofilament (n) profiles and

a microtubule (mt). Note the faintly resolved sidearms extending from the

neurofilaments. The larger structure to the left of the box is a mitochondrion.


carboxyl terminus of NfH protein with b galactosidase; the fusion protein was

sequestrated in perikaryal precipitates as large filamentous aggregates without

export of neurofilaments into the axons. Their axons, devoid of neurofilaments,

were reduced in caliber but did not degenerate. Microtubules and other axoplas-

mic contents instead provided the internal lattice-work required to maintain

structural integrity. Similar atrophic axons were encountered in neurofilament-

deficient quails with a null mutation of NfL [757].

Neurofilaments can be resolved at the ultrastructural level by electron

microscopy. Individual triplet polymers are identified as dots regularly spaced

through the transverse area of the axon with small sidearms (Figure 2.9). Large

myelinated axons may contain hundreds of individual neurofilament profiles

and, while small myelinated and unmyelinated axons normally contain some

neurofilament profiles, their overall number is much smaller. By immunohisto-

chemistry, using neurofilament antibodies, their expression may be minimal,

but it is incorrect to label these axons as lacking neurofilament. The term

“neurofilament poor” or “neurofilament lacking,” when referring to smaller

caliber neurons or axons is therefore technically incorrect.

Prior to the use of immunohistochemical methods, axons were regularly

labeled with metal stains (e.g., Bielschowsky or Bodian silver-based stains) that

had affinity to neurofilament. When carried out by an experienced technician,

a silver-impregnated axon profile can be beautifully resolved. Staining of collagen

and other nerve constituents, however, can sometimes make their interpretation

difficult.

Peripherin is a Class IIIF intermediate filament found in small sensory and

autonomic neurons. The role of a-internexin is less well established. Its mRNA

is only expressed at low levels in intact motor neurons [454] and it may not be

required for the growth in caliber of axons [387].

Interestingly, neither loss of any of the Nf subunits, peripherin, nor a-inter-

nexin is associated with progressive motor neuron degeneration. Most contribute

to the radial caliber of axons and there is axon atrophy when they are knocked

out but none are essential for axon outgrowth [375]. Mice lacking NF-L or NF-M

do have fewer numbers of axons. Alternatively, mice with overexpression or

mutations of some intermediate filaments, such as peripherin, may have alter-

ations in stoichiometry that do lead to a progressive degenerative phenotype [375].

Microtubules consist of polymerized tubulin subunits recognized ultrastructu-

rally as 20–26nM diameter circular profiles (Figure 2.9). They are oriented

longitudinally along the axon and, as the axon diameter decreases, their packing

density increases. This contrasts with the decline in neurofilament numbers as

the axon radial diameter decreases. Most microtubules have approximately

13 subunits in transverse section with a clear center sometimes including

21Axon ultrastructure

a small central dot. Two tubulin subunits, termed a and b, each approximately

50kD in size, form a dimer in the soluble form, but link in the polymer to form a

spiral lattice of subunits. Subtypes of tubulin subunits are found in nerve. Mice,

for example, express a1, aII and bII, bIII, and bIVa [367]. Nerve motor and sensory

axons transport bII and bIII tubulin and incorporate them into microtubules

[205]. a1 tubulin is also incorporated into regenerating axons [472]. The use of

antibodies to bIII tubulin offers a sensitive probe to detect axons in vitro or in vivo.

Polymerization, or addition of tubulin subunits occurs at the “þ” end, or distal

end, of the microtubule and depolymerization or removal of subunits occurs at its

“�” end. Microtubular assembly, in turn, is facilitated by microtubule-associated

proteins (MAPs). In ultrastructure, MAPs can be resolved as “fuzzy”-appearing

extensions from the microtubule. Microtubules associate with the MAP kinesin, a

“motor” to form themachinery for rapid anterograde axoplasmic transport. Antero-

grade axoplasmic transport involves movement of vesicles and approximates

400mm/day. Retrograde axoplasmic transport is approximately 200–300mm/day

and its associated MAP motor is dynein. Slow axoplasmic transport does not rely

on microtubules and is timed at approximately 0.2–2.5mm/day. Slow transport

is further divided into types a and b and it transports structural proteins like

neurofilaments and tubulin. Other MAPs are considered in Chapter 5 in relation-

ship to growth cone dynamics.

Ribosomal periaxoplasmic plaque domains, identified in rabbit and rat

lumbar spinal nerves, are restricted ribosomal domains that may account for

localized protein synthesis in axons [352]. These are narrow (2 micron) elongated

(10 micron) sites randomly distributed and found variably around the periphery

of the axoplasm near the axon–myelin border.

Schwann cell ultrastructure

SCs are recognized in vivo at the ultrastructural level by their close

association with axons and their near invariable investment of basement mem-

brane. By electron microscopy, their nucleus is described as pale without nuclear

condensations or irregularities. Most have pale cytoplasm with a minority (less

than 5%) having dark cytoplasm [477] and most do not have prominent organ-

elles. Reich granules are lamellar, one micron long, metachromatic cytoplasmic

inclusions, likely lysosomal and are found only in SCs associated with myeli-

nated axons [43]. The basement membrane of SCs is easily recognized under the

EM as irregular “fuzzy” material just outside of the plasmalemma. It can be

divided into two layers known as lamina densa (dense region) and an inner

lamina lucida (clear space) [43] and it is composed of collagen Type IV, laminin,

and fibronectin. Single SCs associated with myelinated axons are distinct units.


SCs associated with unmyelinated axons normally ensheath several in a Remak

bundle, as previously discussed. A single Remak SC invagination is usually

associated with only one, or at most two unmyelinated axons.

Myelin sheaths

Myelin sheaths consist of lipid membrane material, synthesized by SCs

in the PNS and oligodendrocytes in the CNS. PNS myelination involves the

spiraling of SC membranes around axons to form sheaths of between 60 and

120 layers, or lamellae. The number of lamellae is a function of the size of the

axon. The G ratio is the ratio of the axon (alone) diameter to that of the

myelinated fiber diameter (including myelin). It is 0.65 in a normal human sural

nerve and is independent of axon size. Thus, larger caliber axons have more

myelin lamellae and an overall thicker myelin sheath than smaller axons in

keeping with a normal G ratio. Axons with thin myelin during remyelination or

regeneration have a larger G ratio, approaching 1.0.

When myelin is formed from spiral loops of SCs, the cytoplasm is largely

squeezed out between their membrane surfaces. This results in distinct lines

within each loop of myelin. The intraperiod line is formed from the apposition of

the extracellular surfaces of the SC membrane prior to compaction. The major

dense line represents the apposition of their cytoplasmic surfaces. In unfixed

tissue, the periodicity of myelin in approximately 18nm [43]. The inner mesaxon

is the portion of retained SC cytoplasm and membranes attached to the myelin

sheath found immediately adjacent to the axon. It is separated from the axon

proper by the periaxonal space. The outer mesaxon is that portion of the

SC cytoplasm and membrane contacting the outer end of the myelin sheath.

Schmidt–Lanterman clefts or incisures are separations of the compact myelin

lamellae by SC cytoplasm that form elegant spiral connections between the

inside and outside of the myelin. They are thought to represent portals for

transport between the inner adaxonal SC cytoplasm and the outer abaxonal

SC cytoplasm.

The node of Ranvier is a highly specialized structure that constitutes 2%–3% of

the length of an internode. The axon is constricted in diameter at nodes of

Ranvier with reductions in transverse area of 75%–90% and rises in neurofilament

packing density [43] (Figure 2.7). The borders of the node are demarcated by

bulbous processes of SC cytoplasm rich in mitochondria, so-called paranodal

bulbs that form nodal “collars” [43]. The SC processes are formed from the

abaxonal cytoplasm and are associated with basement membranes. Nodal collars

send microvilli, forming a brush border within the node. The node itself contains

matrix material, one component of which, a chondroitin sulphate proteoglycan

23Myelin sheaths

(CSPG; see C10), may serve to inhibit sprouting in stable nonregenerating axons

[43]. Paranodes are the segments of the myelinated axon just outside the node of

Ranvier. Paranodes thus contain loops of the SC known as terminal cytoplasmic

spirals that attach to the axolemma with septate-like junctions [43]. The terminal

SC spirals and associated paranodal axon are known as the glial–axon junction.

The juxtaparanode is found lateral to each paranodal segment. An additional

term used is paranode–node–paranode (PNP) referring to node and both of its

neighboring paranodal segments. There is a complex and intimate association

between the axon and adaxonal SC in the PNP region with interdigitating

processes known as the “axon–SC network” [43].

Overall, myelin is about 70% lipid but also contains an important complement

of protein, comprising the remaining 30% [209]. The lipid contains cholelsterol

and glycosphingolipids, whereas proteins of the myelin sheath in peripheral

nerves have strict localization [661]. P0 (MPZ), the most common myelin protein,

is an adhesion molecule that plays a role during developmental myelin compac-

tion and persists in mature compact myelin. PMP 22 (peripheral myelin protein 22)

is a membrane protein found in compact myelin. It has complex roles that

not only influence myelination but also cellular survival and proliferation.

The PMP 22 gene is abnormal in Charcot–Marie–Tooth disease (CMT) Type 1a,

an inherited demyelinating polyneuropathy. MBP (myelin basic protein, also

called P1) is a cytoplasmic, membrane-associated protein in compact myelin that

also aids in compaction during development. P2 protein is found in the major

dense line of compact myelin. MAG (myelin-associated glycoprotein) is an adhe-

sion membrane molecule found in noncompacted myelin that helps to maintain

myelin and axon integrity. It is also a ligand of the Nogo receptor that collapses

growth cones and inhibits regeneration. Cx32 (connexin 32) is a tunnel protein

of gap junctions found in noncompacted myelin that is involved in intercellular

communication. E-cadherin is a membrane adhesion molecule found in adhe-

rens junctions of noncompacted myelin in the mesaxons. Neurofascin 155 is

a membrane adhesion molecule of SCs found on the adaxonal side (facing

the axon) of paranodal loops. TAG-1 is a SC adhesion molecule found in juxtapar-

anodal myelin. MAL (myelin and lymphocyte protein) is a tetraspan proteolipid

found in compact myelin and with unmyelinated fibers. Periaxin is a

SC cytoskeletal-associated protein found on the abaxonal compartment (facing

the extracellular matrix) of myelinated fibers.

Molecular structure of myelinated axons

Axonal and SC proteins comprise a complex and highly structured

pattern in the myelinated fibers of the peripheral nervous system. A variety


of inherited (e.g., CMT) and acquired (e.g., autoimmune neuropathy with

autoantibodies directed against MAG) polyneuropathies involve alterations of

proteins required for axon–SC interaction. By disrupting a single element of

the interaction, serious consequences (demyelination, axon damage) for the

peripheral nerve ensue.

The node of Ranvier is the excitatory nexus of the myelinated peripheral

nerve. To convey action potentials through saltatory conduction (not reviewed

here), these nodes have a concentration of membrane spanning sodium chan-

nels. Specific subtypes of Naþ channels identified at mammalian nodes of

Ranvier are Nav 1.6, the commonest subtype but also Nav 1.2, Nav 1.8 (SNS/PN3),

and Nav 1.9 (PN1). The latter two channels are classified as tetrodotoxin (TTX)

resistant because their currents are not blocked by TTX. Nav 1.7 is present in

both sensory and sympathetic neurons [748]. Nav 1.3 (Type III) is a channel that is

re-expressed after injury. All of the subtypes are voltage gated and share a

24-membrane spanning architecture through four distinct a subunits. b subunits

are auxiliary components. AnkyrinG proteins anchor the sodium channels at

nodes and interact with the cytoplasmic proteins neurofascin, Nr-CAM, and

spectrin. The paranode axonal membrane contains heterodimers of contactin

and Caspr proteins and the associated SC membrane contains neurofascin 155,

all of which are components of the septate-like junctions mentioned above [604].

Further lateral to the paranode is the juxtaparanode, an area that contains

potassium channels, especially Kv1.1 and Kv1.2 linked to Caspr2. The Kv potas-

sium channels, hidden beneath layers of myelin, likely act as rectifiers that

can dampen axon excitability. Kv1.5 channels are also found on SCs, in their

abaxonal portion where they may serve a buffering function during axonal

conduction. Disruption of the normal paranode and juxtaparanode may inappro-

priately expose axonal channels and interfere with normal myelinated axon

conduction. The internode of the axon, located between nodes or more properly

between the flanking juxtaparanodes, contains fewer proteins. As may be evident,

the complex molecular anatomy of the node of Ranvier is critically important to

axon excitability and impulse transmission and to axon–SC interactions. Readers

are invited to detailed depictions of the node found elsewhere [23,604].

Motor neurons and the anterior horn of the spinal cord

Motor neurons consist of two populations: larger caliber myelinated

a motor axons that signal voluntary control of skeletal or striated muscles and

smaller caliber myelinated g motor axons that innervate muscle spindles (intra-

fusal muscle fibers). Alpha motor axons have conduction velocities of 40m/s

or higher. By definition, motor neuron perikarya are placed within the Rexed

25Motor neurons and the anterior horn of the spinal cord

lamina (laminae that define the cytoarchitecture of the spinal cord) X that is

found in the anterior or ventral horn of the gray matter of the spinal cord.

Within this lamina, they are intermixed with numerous smaller interneurons.

Motor neurons are large, polygonal and contain abundant rough endoplasmic

reticululm with extensive dendritic arbours in the the gray matter of their

ventral horn (Figure 2.10). Motor neurons innervating appendicular, or limb

muscles are found more laterally in the anterior horn, whereas those innervat-

ing axial or trunk muscles occupy more medial portions. Since more motor

neurons are present in the cervical and lumbar regions of the spinal cord to

supply the upper and lower limbs, respectively, there are cervical and lumbar

enlargements at these levels. From the anterior horns, axons exit the spinal cord

through the ventral roots, then join the dorsal sensory roots to form the mixed

spinal nerve. Autonomic axons from the intermediolateral cell columns also exit

the spinal cord through the ventral roots.

A motor unit refers to a single motor neuron that has branches to a number of

individual muscle fibers. The size of the motor unit can vary substantially among

different muscles but once activated, all of its muscle fibers normally fire

synchronously. The physiological properties of the muscle fibers within a given

Figure 2.10 An illustration of an anterior horn a motor neuron. Note its triangular

shape and its content of dark Nissl bodies. The oil immersion image is from a semithin

section, stained with toluidine blue from a mouse spinal cord. (Image taken by

Noor Ramji, Zochodne laboratory.)


motor unit are dictated by the properties of the axon that innervate them.

In some muscles, such as extraocular muscles, motor units may be small inner-

vating only five muscle fibers (also known as their innervation ratio). In others

they are large and may innervate 2000 or more muscle fibers, as in the human

gastrocnemius muscle [145]. The size of the motor unit has a direct bearing on

the precision required of the muscle it innervates. For example, a large limb

muscle is required to power activities such as walking that requires synchroni-

zed firing of large numbers of muscle fibers. Their motor units are therefore

larger. In the eye, however, very small precise movements are required to

maintain fixation of the fovea on a visual target. Fine gradations of control

and smaller motor units are required.

Sensory neurons and ganglia

Sensory neuron cell bodies, or perikarya, reside in cranial and spinal

ganglia. When identified in humans at surgery or autopsy, ganglia are observed

in close relationship with nerve roots as an enlargement associated with fat

tissue. A distinct pedicle composed of entering and exiting axons can sometimes

be identified (Figure 2.11). Ganglia associated with lumbar and cervical roots are

larger and contain more neurons than those in the thoracic spine. Surrounded

by a specialized capsule, large and small neurons are freely intermixed and each

is surrounded by perineuronal satellite cells. Satellite cells have little cytoplasm

and their flattened nuclei closely appose neurons. Neuron perikarya and their

satellite cells characteristically line the cortex of the ganglia just beneath the

capsule, whereas the center of the ganglion includes some neurons but is

particularly invested by axons and their associated SCs. Blood vessels supplying

dorsal root ganglia arise from adjacent radicular (root) branches supplying the

nerve roots and spinal cord and indirectly by connections from the anterior

spinal artery [3]. These vessels supply local intrinsic blood flow to ganglia that

is higher (30–40ml/100g per min) than that of the peripheral nerve trunk and

has some degree of autoregulation [792] (see Chapter 7). The blood–ganglion

barrier is also less robust than that of the nerve [24]. Thus, proteins access ganglia

much more easily than nerves, a feature that may have important implications

for neuropathies related to blood borne antibodies and toxins.

Like their axons in the peripheral nerve trunk, neurons in DRGs are distrib-

uted into two major size categories. Classically, these are divided into large

light neurons with Nissl substance (rough endoplasmic reticulum) separated by

neurofilaments and small dark neurons with more concentrated Nissl substance

and a lower content of neurofilaments (Figure 2.12). Neurofilaments in the cell

body or perikaryon similarly form a lattice network that may anchor the nucleus

27Sensory neurons and ganglia

into a central position within the cell. Thus when staining DRG neurons by

neurofilament immunohistochemistry, the larger neurons may be preferentially

stained because of their higher neurofilament content (Figure 2.13). The size of

DRG neurons range from 10–100 microns in diameter or a cross-sectional area

of under 500mm2 for small neurons of unmyelinated fibers or Ad fibers and

above this for Aab fibers [379].

Figure 2.11 An illustration of a primary sensory neuron that resides in a dorsal

root ganglion. The neuron is described as “pseudounipolar” with a single branch from

the perikaryon (cell body) that then divides into central and peripheral branches.

The perikaryon is surrounded by closely apposed perineuronal satellite cells.

(Illustration by Scott Rogers.) See color plate section.

Figure 2.12 A photomicrograph of a portion of a lumbar dorsal root ganglion

(DRG) illustrating peripherally placed large and small sensory neurons and more

centrally placed myelinated axons. The image is a semithin section from a rat L5 DRG

stained with toluidine blue.


The types of sensory neurons that populate dorsal root ganglia can be classified

in several further ways (see review [379]). Some peptides are distributed selectively

to small neurons associated with C and Ad axons: ET1 (endothelin-1), galanin,

nociceptin, somatostatin, SP (substance P), PACAP (pituitary adenylate cyclase-

activating polypeptide), VIP (vasoactive intestinal polypeptide); or to small- and

medium-sized neurons: CGRP (calcitonin gene-related peptide) (Figure 2.14).

Receptors expressed by small neurons include: GLUR1 (glutamate receptor),

GLUR5, AMPA, Kainate, B1 (bradykinin receptor), B2, H1 (histamine receptor),

EGF receptor (epidermal growth factor receptor), FGF2 (fibroblast growth factor 2

receptor), NK1 (neurokinin-1 receptor), SSTR2a (somatostatin receptor), P2X3

(purinergic receptor), GAL2 (galanin receptor), ETA (endothelin A receptor),

GFRa3 (GDNF family receptor); by small and medium neurons: TrkA, p75, GFRa1

(GDNF family receptor), GFRa2, RET (GDNF family coreceptor), MOR (m opioid

receptor), DOR (d opioid receptor), KOR (k opioid receptor), ORL1 (opioid-like

receptor 1); by large neurons: TRPV2 (transient receptor potential family), CB1

(cannabanoid 1 receptor), GAL1, GLUR2/3, GM1 (GM1 ganglioside receptor), P2Y1

(purinergic receptor), TrkC. Ion channels expressed by small neurons include: BK

(bradykinin), KCa (calcium-activated potassium channel), L and N type Ca2þ chan-

nels (calcium channels), Kv1.4 (potassium channel), Nav1.9 (sodium channel), Nab3,

TRPA1, TRPM8, TRPV1, TRPV4, TRPV3; by small and medium neurons: Nav1.7, 1.8,

Figure 2.13 An image of DRG sensory neurons labeled by immunohistochemistry with

an antibody directed against the heavy subunit of neurofilament (red). The arrowhead

points to a smaller neuron with less prominent neurofilament staining (sometimes

erroneously called “neurofilament negative”). The right (green) panel is the same

section with transmitted light. The inset shows a single large neuron with a

“unipolar” process (arrow) emerging from the perikaryon. Note also that the

neurofilament label shows a patchy and complex network within the perikaryon

that is thought to determine overall neuronal shape (Bar ¼ 20 microns).

(Image taken by Chu Cheng, Zochodne laboratory.) See color plate section.

29Sensory neurons and ganglia

P2X2/3; by large neurons: Nab1.1, 2,1, Kv1.1, 1,2, Kvb2.1, HCN1,2 (hyperpolarization-

activated cyclic-nucleotide gated, time- and voltage-dependent nonselective cation

channel). This list is not complete, however, as there are cytokines and enzymes

also distributed differently among these neurons. Similarly, they can be divided

into their sensory target receptive properties and their projections.

The expression of high affinity receptors for neurotrophin growth factors [325] is

particularly relevant here. Neurons expressing TrkA, the receptor for NGF are

preferentially found on small caliber neurons that subserve nociception. These

receptors are found on 41% of lumbar DRG neurons. TrkB, the cognate receptor

for the BDNF neurotrophin family member is found on 33%, and TrkC, the

receptor for NT-3 on 43%. TrkB is found on small- to medium-sized neurons

and Trk C on large neurons. P75, formerly known as the low affinity neurotro-

phin receptor, and capable of binding all family members, is found in 79% of

DRG neurons. There is, however, extensive colocalization of Trk receptors and

trilocalization in a small percentage. Non Trk bearing neurons are the IB-4

population that express the GDNF receptors GFRa-1 and RET and express the

Figure 2.14 An image of a rat lumbar DRG labeled by immunohistochemistry with

an antibody directed against the peptide CGRP. Approximately 45% of DRG sensory

neurons are medium- and small-sized CGRP positive. Note the emerging axons

from the DRG also exhibit CGRP immunoreactivity. (Image taken by Chu Cheng,

Zochodne laboratory.)


P2X3 purinoreceptors (ATP sensitive) [59]. The interaction of growth factors with

peripheral neurons is addressed in Chapter 9.

Autonomic, enteric ganglia

Peripheral autonomic neurons consist of sympathetic and parasympa-

thetic neurons. The size of a given autonomic ganglion depends on the extent of

its target tissue. Sympathetic ganglia are found as chain spinal (also called

paravertebral) ganglia or along nerves called satellite ganglia. In humans, the

paravertebral ganglia extend from C8 to S5 bilaterally along the spinal column.

The ganglia receive preganglionic axons from the spinal cord (also known as

white rami communicans (C8-L2, myelinated)) that are branches from the ventral

roots. Postganglionic axons, unmyelinated, leave the ganglia as gray rami com-

municans and rejoin mixed spinal nerves. They eventually innervate blood

vessels, sweat glands, pilomotor fibers, or join plexi that innervate most internal

organs. The ganglia below L2 do not receive white rami communicans but receive

preganglionic fibers from higher levels and send out gray rami. The sacral chains

are fused. The C8 and T1,2 roots send preganglionic fibers up the sympathetic

chain along the carotid artery to supply the stellate (inferior cervical), middle,

and superior cervical ganglia. Postganglionic axons either travel further along

the chain or branch off along the carotid arteries as they bifurcate and they

supply the head and neck. Other preganglionic axons from lower levels join

prevertebral ganglia known as the celiac, superior mesenteric, and inferior

mesenteric ganglia.

The parasympathetic fibers are craniosacral in their distribution in humans.

This means that they originate in cranial nuclei: third cranial nerve (Edinger–

Westphal nucleus), sending preganglionic axons to the ciliary ganglion that, in

turn, supply postganglionic axons to the eye; seventh cranial nerve that sends

pregangionic axons to the sphenopalatine and submandibular ganglia to supply

tearing and salivation; the ninth cranial nerve supplying preganglionic axons to

the otic ganglia; the vagus, or tenth cranial nerve sending long preganglionic

axons to the pulmonary tree, heart, liver, stomach, intestines, and kidney into

ganglia associated with these organs that then project small postganglionic

axons to them. The parasympathetic fibers also originate from S2, 3, and 4 sending

preganglionic axons in the pelvic splanchnic nerves to supply ganglia associated

with the colon and urogenital systems.

Autonomic ganglion neurons have a similar size distribution as sensory

neurons, but they are multipolar instead with a number of dendrites and a

single axon. Their intertwining dendrites can form “dendritic glomeruli or

nests” [363] forming a complex anatomy that includes both axodendritic and

31Autonomic, enteric ganglia

axosomatic synapses. In addition, some sensory neurons are found in autonomic

ganglia. Sympathetic neurons express the peptides NPY (neuropeptide Y), soma-

tostatin, VIP, and enkephalin. In parasympathetic ganglia VIP and CGRP are

identified. Most sympathetic ganglion neurons are noradrenergic and contain

tyrosine hydroxylase (Th, a catecholamine synthesizing enzyme). Most parasym-

pathetic ganglion neurons are cholinergic, and contain choline acetyltransferase

(ChAT, an acetylcholine synthesizing enzyme). While this arrangement exists

for most sympathetic axons, some that project to the head and neck vessels or

to sweat glands of the skin instead are cholinergic. Unmyelinated postganglionic

autonomic neurons contact their endorgans through “neuroeffector junctions.”

Unlike classical synapses, these junctions lack presynaptic concentrations of

neurotransmitters, have wider and more variable clefts, and do not have a

classical postsynaptic structure.

The enteric nervous system consists of afferent, interneuronal, and motor

axons and extends from the pharyngoeseophageal junction to the internal anal

spincter in man. It includes the myenteric (Auerbach’s) plexus that is largely

motor in function, found between the longitudinal and circular smooth muscle

layers of the intestine, and the submucous (Meissner’s) plexus that has a role in

regulation of secretion in the submucosal layer. The enteric nervous system thus

has close interactions with sensory, sympathetic, and parasympathetic neurons

(see review [616]).

Innervation of the skin

The innervation of the skin involves a system of overlapping sensory

transducers placed at varying depths within the epidermis and dermis. Sensa-

tions of pressure, light touch, vibration, pain (or nociception), heat, and cold are

transmitted to nerves, the spinal cord, and the brain. Pressure and light touch

might be more rigorously classified as “cutaneous displacement” [741]. Slowly

adapting (SA) units include type SAI that consists of myelinated axons innervat-

ing Merkel touch complexes. Merkel cells are recognized by their indented nuclei

and cytoplasmic granules and can form into dome-like structures [33]. They are

found both in the basal layer of the epidermis of both hairy and hairless

(glabrous) skin [152] whereas in rodents, they are also associated with whisker

vibrissae [152]. They have small receptor fields. SAII innervate Ruffini endbulbs

deeper in the dermis and have wider receptive fields. These structures measure

1mm by 30 microns and consist of a branched array of terminal myelinated axon

branches [33]. Cutaneous velocity detecting units are, in contrast, rapidly

adapting and include receptors known as G2 hair follicle and T hair follicle

receptors, D hair follicle receptors, F field receptors, and C mechanoreptors.


For example, G2 and T receptors respond to the velocity of hair movements, D are

more sensitive to slight movements in smaller, down hairs, and F, field receptors,

require movements of groups of hairs. C unmyelinated skin indentation recep-

tors respond to the velocity of slow skin movements. In glabrous (nonhairy) skin,

the rapidly adapting (FAI; fast adapting I) receptor with a small receptor field is

the Meissner’s corpuscle in the dermal ridges. Meissner’s corpuscles are 80 by 30

microns in size, oriented perpendicular to the surface of the skin and are in the

form of a corpuscle supplied by 3–4 myelinated axons in a knot infiltrated with

SCs and a collagenous capsule [33]. There are also transient cutaneous detectors

that include G1 hair follicle receptors and Pacinian corpuscles. Pacinian cor-

puscles are formed at the ending of myelinated fibers that are inserted into

layers of flattened cells in an overall ellipsoidal shape of 1mm by 0.7mm [33].

Pacinian corpuscles are very sensitive to movement or vibration, especially in the

range of 60–300Hz and are found deeper in the dermis to provide a wide receptor

field. Some of these receptors generate ongoing activity at ‘‘rest.’’

Free nerve endings of unmyelinated C fibers that end in the epidermis and

Ad fibers in the dermis convey pain (nociception) and thermal sensation

(Figure 2.15). Overall, nociceptors have been classified as mechanical nociceptors

Figure 2.15 An image of the skin of a normal mouse footpad labeled by

immunohistochemistry with an antibody directed against the axon marker PGP 9.5

(green). Note the dense ramification of tortuous axons up from the dermal plexus

into the epidermis. The epidermis is also covered by thick callus. (Image taken by

James Kennedy, Zochodne laboratory and reproduced with permission from [335].)

See color plate section.

33Innervation of the skin

(Ad fibers), polymodal (sensitive to several modes of pain) nociceptors (C fibers),

other mechanical nociceptors (C), mechanoheat nociceptors (Ad), and cold noci-

ceptors (Ad or C fibers). Chemoreceptors may overlap with these groups. There

are also thought to be “silent” nociceptors only activated by inflammation and

injury. The biology of sensory transduction, however, has progressed well beyond

these classical concepts and the classifications do vary depending on the refer-

ence (see review by [419]). I will mention just a few of the new molecules linked

to sensory transduction. In particular, the TRP (transient receptor potential)

molecules have assumed a central role in this evolving story. Heat transduction

is mediated by TRPV1, the classical capsaicin receptor or vanilloid (“V” in TRPV)

receptor and also by TRPV2, 3 and 4, depending on the intensity (temperature).

In contrast, TRPM8 was discovered as a cold transducer, or menthol (“M” in

TRPM) receptor. Another receptor TRPA1 also contributes to cold sensitivity,

but it is also important for acute mechanical transduction. TRPN and DEG/ENaC

channels may mediate touch and proprioceptive stimuli. It seems likely that

cutaneous fibers may express more than one class of TRP in a single fiber

indicating a multiplicity of function depending on the exact stimulus.

The skin is also innervated by sympathetic axons that innervate sweat glands

(cholinergic) in the dermis and arrector pili muscles (adrenergic) of the dermis

(that cause “goose flesh” when activated; also called pilomotor). Sweat glands are

divided into eccrine glands, the most widely distributed types in skin, and

apocrine glands that are associated with hair follicles. In humans apocrine glands

are localized to axillae, perineum, and nipple areolae. Sweat glands are innervated

by sympathetic cholinergic axons (receptors called M3). The gland includes a

secretory coil where sweat is produced and secretory duct that expels sweat while

reabsorbing NaCl. The entire apparatus is contained within a sweat gland capsule.

Innervation of muscle, joints, viscera

Muscles have a rich sensory nerve supply [741]. Muscle spindles are

specialized fibers (also called intrafusal) that possess sensory organs supplied

by large myelinated axons sensitive to dynamic movements (Group Ia) and by

intermediate-sized myelinated axons sensitive to muscle length (static; Group II).

Intrafusal muscles (spindles) are also classified into nuclear bag fibers (innervated

by Group I primarily) and nuclear chain fibers (innervated by both Group I and II).

These muscle fibers also receive input from specialized smaller caliber motor

axons known as g motor neurons. By controlling the activation of the intrafusal

muscle fiber, g motor axons can thus influence its tension (it will also be under

stretch because it is connected in parallel with a large number of extrinsic,

normal or “extrafusal” muscle fibers) and hence influence the sensitivity of


the receptors. Golgi tendon organs (innervated by Group Ib afferents) are located

in the connective tissue of muscle tendons and they are high threshold receptors

to passive muscle stretch. Up to 75% of muscle innervation, however, is from

free nerve endings, usually unmyelinated axons that are mainly nociceptive,

or sometimes thermal sensitive. There are also Ad (Group III) fibers in muscle

that supply information about mechanical stimulation, pressure, chemical

stimulation, stretch, and nociception.

Position sensibility is mediated by muscle and tendon stretch, but also

by joint afferents. As in the skin, these include Ruffini slowly adapting joint

receptors, transient-sensitive Pacinian corpuscles and joint Ad and C nociceptors.

In the viscera, there is innervation by Pacinian corpuscles in the mesentery

and connective tissue, some mechanoreceptors and prominent nociceptors

particularly involving the surface of viscera and mucosal linings.

Table 2.1 Cutaneous Innervation*

Name Location Cutaneous field Function

Mechanoreceptors

SAI (Merkel) Derm/epid jct Small Displacement

SAII (Ruffini) Dermis Medium-large Displacement

FAI (Meissner’s) and

?Krause endbulb

Dermis Small Velocity

Pacinian Dermis, other areas Large Transients, vibration

C Epidermis and

dermis

Unknown Velocity, displacement

Hair mechanoreceptors

G1 Large hairs Single follicle Transients

G2,T Large hairs Single follicle Velocity

D Small down hairs Single follicle Velocity, low threshold

Field Hair and ?skin Multiple hairs, skin Velocity, displacement

Nociceptors

Ad mechanical Dermis/?epidermis Small Damage

C polymodal Epidermis Small Damage, heat,

chemical, other

C mechanical Epidermis Small Damage

Ad mechanoheat Dermis/?epidermis Small Damage, heat

Ad and C cold Dermis/?epidermis Small Cold

Note:

*Adapted and simplified from [741].

35Innervation of muscle, joints, viscera

Neuromuscular junction

Neuromuscular junctions are specialized synapses that connect motor

axons to muscles. They consist of the motor terminal, the junctional cleft, and

the postsynaptic muscle endplate, also known as the motor point (Figure 2.16).

Each has a specialized structure. The motor axon terminals lose their myelin

sheaths but become closely associated with specialized SCs known as presynaptic

SCs. The terminals splay out and become filled with synaptic vesicles approxi-

maly 50nm in diameter, mitochondria and tubules of smooth endoplasmic

reticulum. Finally, they form a synaptic gutter, an indentation into the muscle

known as the primary fold. There is also loss of the blood–nerve barrier at

this site.

Muscle fibers innervated by a given motor axon are not usually grouped

together but intermingle with those of other motor units. The contractile prop-

erties of muscle fibers are dictated by the motor axon innervating them, i.e., slow

sustained contraction oxidative (Type I) fibers compared with rapid fast twitch

glycolytic (Type II) fibers. At the prejunctional terminal, there are P/Q-type

voltage gated Ca2þ channels (VGCCs) that depolarize the terminal and promote

release of acetylcholine from synaptic vesicles. Voltage-gated potassium channels

Figure 2.16 Images of neuromuscular junctions from the mouse hindlimb labeled

by immunohistochemistry with an antibody directed against neurofilament

(NF200, green) identifying terminal motor branches. The junction itself is labeled with

a bungarotoxin (a-bTx, red) that binds the presynaptic terminals. (Image taken

by Noor Ramji, Zochodne laboratory and reproduced with permission from [571].)



(VGKCs) help to restore the membrane potential of the presynaptic space after

depolarization and calcium influx. Choline acetyltransferase (ChAT), the enzyme

necessary to synthesize the neurotransmitter acetylcholine (Ach) is found in the

presynaptic terminal.

The Ach vesicles interact with electron dense specialized zones of the axonal

membrane that are called active zones. By ultrastructure, active release zones

resemble a thickened bar associated with parallel arrays of synaptic vesicles that

have a close association with the VGCCs. Spontaneous release of single quanta

(packets or vesicles) of Ach generate discharges in the muscle endplate zone

known as MEPPs, or miniature endplate potentials. An action potential in the

presynaptic motor nerve causes influx of calcium, local depolarization, and

massive vesicular release of Ach into the synaptic cleft. Ach then diffuses to

the postjunctional complex. In the cleft it binds to postjunctional Ach receptors

(AchR) or is degraded by acetylcholinesterase found with basement membrane in

the cleft. In the muscle endplate, Ach binding of receptors (AchR) generates an

endplate potential which, if above threshold, causes a muscle action potential.

The neuromuscular junction is approximately 60–100nm in width and it is lined

by a basement membrane that closely follows the junctional folds. The postjunc-

tional complex consists of a folded membrane complex (junctional folds or

secondary folds) on the apices of which sit acetylcholine receptors. These consist

of two a subunits and one each of d, b, and g subunits that form an ion pore.

AchR are also closely associated with the proteins MuSK (muscle specific kinase)

and rapsyn (receptor aggregating protein at the synapse). When ligated by Ach,

the ion pore opens to allow entry of sodium, potassium, and other small ions

with subsequent depolarization. Depolarization, in turn, turns on the actin–

myosin contractile apparatus that is the basis of muscle contraction.

Summary

Peripheral neurons connect the central nervous system to muscles and

sensory organs. In doing so, they traverse diverse territories and incorporate

several types of structures. Far from being uniform, peripheral nerve trunks

contain a mixture of myelinated motor axons, myelinated and unmyelinated

sensory axons, and unmyelinated autonomic axons. Each originates in different

parts of the nervous system: ventral horn of spinal cord for motor neurons,

ganglia for sensory and autonomic neurons. Each also has separate destinations

but while en route they share pathways in nerve trunks with unique anatomical

and physiological characteristics. An appreciation of the diverse complexity of

peripheral neurons is essential to understand how they respond to injury during

regeneration.

37Summary

Suggested reading

Berchthold, C.-H. Fraher, J. P., King, R. H. M., & Rydmark, M. (2005). Microscopic

anatomy of the peripheral nervous system. In Peripheral Neuropathy, Chapter 3, ed.

P. J. Dyck & P. K. Thomas, 4th edn. Philadelphia: Elsevier Saunders [43].

King, R. H. M. (1999). Atlas of Peripheral Nerve Pathology. London: Arnold Press [346].

Lawson, S. N. (2005). The peripheral sensory nervous system: dorsal root ganglion

neurons. In Peripheral Neuropathy, Chapter 8, ed. P. J. Dyck & P. K. Thomas, 4th edn.

Philadelphia: Elsevier Saunders [379].

Peters, A., Palay, S. L., & Webster, H. deF. (1991). The Fine Structure of the Nervous System,

3rd edn. Oxford, UK: Oxford University Press [554].

Schroder, J. M. (2001). Pathology of Peripheral Nerves. Berlin: Springer [608].

Shaw, G. (1998). Neurofilaments. Georgetown: Springer [619].

Suter, U. & Martini, R. (2005). Myelination. In Peripheral Neuropathy, Chapter 19, ed.

P. J. Dyck & P. K. Thomas, 4th edn. Philadelphia: Elsevier Saunders [661].

Willis, W. D. & Coggeshall, R. E. (2004). Sensory Mechanisms of the Spinal Cord, 3rd edn.

New York: Plenum [741,743].


3

Injuries to peripheral nerves

Peripheral nerve injuries are common, disabling and difficult to treat.

These injuries arise because nerve trunk anatomical pathways expose them to a

variety of injury types, including both blunt and penetrating lesions. The injury

type, in turn, has a major bearing on how nerves respond. There are unique and

challenging attributes of nerve trauma biology that set the stage for subsequent

regenerative activity. These attributes are examined in this chapter.

Overall considerations

The type of injury experienced by a peripheral nerve determines

how regeneration might proceed. Several classifications exist and are given here.

The simplest classification is provided by Seddon, who grouped injuries into

neurapraxia, axonotmesis, and neurotmesis lesions.

Neurapraxia is a lesion that occurs most often from blunt nerve trauma and

involves focal demyelination over several internodes at the injury site. By defini-

tion, there is no associated axonal injury. Since neurapraxic lesions are identi-

fied by the loss of function associated with them, they require that enough

internodes be disrupted to block conduction, or function. For example, a

common type of neurapraxic lesion is the “Saturday night palsy” of the radial

nerve in the arm that develops from prolonged compression of the inner arm by

a chair or the head of a partner (and is often associated with alcohol intoxica-

tion!). Compression, however, need not be prolonged, since it is highly likely

that the mechanical distortion of the nerve is responsible for demyelination.

Ochoa and colleagues, in experimental work, described how myelinated axons

“intussuscept” from undue mechanical stress [521]. Intussusception is the tele-

scoping together of two hollow tubes, as most often described in infants, where

39

a portion of small intestine is forced into a neighboring segment. For the injured

nerve, one internode’s myelin sheath is forced up and over the top of its

neighbor. The distorted myelin disrupts salatatory conduction, unsheathes in-

hibitory potassium channels, and triggers myelin phagocytosis. Thus, following

the injury, local SCs and resident macrophages, followed by ingressing circulat-

ing macrophages, are signaled to remove the abnormal intussuscepted myelin.

Segmental demyelination, or loss of myelin strictly in segments between nodes

of Ranvier then results (Figure 3.1); there may, of course, also be several adjacent

demyelinated internodes. There is acute block of conduction over this segment

with loss of distal function. In the case of “Saturday night palsy,” the patient

awakens with a severe and complete wrist drop from inability to extend the

wrist and fingers, normally the work of the intact radial nerve. There is also

sensory loss in the territory of the radial nerve. The adjacent SCs are not

destroyed by the lesion and begin to proliferate and eventually remyelinate

the demyelinated segment. Since several SCs may now be associated with an

internode previously myelinated by one SC, the repaired internodes may be

shorter in length than the original ones (“shortened internodes”) [268,269].

While remyelination, albeit with shortened internodes, alleviates conduction

block, there may be persistent long-term mild slowing of conduction velocity

through the segment. With the passage of time, internodes likely remodel and

probably lengthen back out towards their original proportions [2,268,269].

The precise mechanism by which SCs regulate internodal length is uncertain.

Axonotmesis is a lesion from a blunt mechanical, chemical, or ischemic lesion

in which axons are interrupted but the connective tissues and epineurium of

the nerve trunk are not (Figure 3.2). Therefore, anatomical continuity between

the stump of the nerve proximal to the injury and the distal stump is main-

tained. Other classifications described below subdivide axonotmesis lesions

depending on the extent of associated connective tissue damage. In axonotmesis

injuries, it is likely that a very rapid axonal signal is generated by the trauma

that initiates irreversible damage to the distal axon. Wallerian degeneration,

discussed below, ensues.

Figure 3.1 An image of a teased myelinated fiber illustrating segmental

demyelination between two nodes of Ranvier (arrows). Demyelination accounts

for neurapraxic nerve injuries. See color plate section.

40 Injuries to peripheral nerves

Neurotmesis refers to a lesion in which the entire nerve trunk including

the connective tissues and epineurium is sundered, thus separating the distal

from the proximal stump (Figure 3.3). This lesion develops from a penetrating

injury, transection from a nearby fracture bony fragment, or from a surgical

mishap. Its severity is heightened because the distal and proximal stumps of

the nerve, already under some tension, commonly retract from one another.

For regeneration to occur, these bridges must be reconnected by new connective

tissue and bridging axons or by surgical apposition. At times, the stumps may

retract to the extent that their ends are difficult to identify surgically, especially

if there is other associated tissue injury. Associated tissue damage undoubtedly

may also play a large role in the repair and regenerative process. For example,

compartment syndromes can occur in injured limbs, following trauma in

which tissue pressure, confined by fascial planes, rises higher than vascular

perfusion pressure. Severe ischemia results. Penetrating injuries also pose a

significant risk for superimposed infection, an important cause of failed healing.

Surgeons repairing nerves therefore must attend to issues of vascular repair,

limb stabilization, nerve decompression, and other critical measures.

Figure 3.2 An illustration of early events following a peripheral nerve trunk crush

(axonotmesis). Distal to the center of the crush zone, Wallerian-like degeneration

begins with breakdown of myelin and axons. (Illustration by Scott Rogers.)


41Overall considerations

Other forms of nerve injury classification include that of Sunderland [660].

(i) First-degree injury, similar to neurapraxia above, is defined as an injury

without structural interruption of the axon or subsequent Wallerian-

like degeneration. The lesion is characterized by local demyelination.

It is possible that temporary focal “stunning” or focal dysfunction of

an axon without demyelination may account for a neurapraxic lesion

but no direct evidence for this following trauma has been described.

Temporary axon conduction block is described, however, after acute

ischemia, discussed in Chapter 7 [314,543].

Figure 3.3 An illustration of early events following a peripheral nerve trunk

transection (neurotmesis). In the distal stump, Wallerian degeneration begins with

breakdown of myelin and axons. (Illustration by Scott Rogers.) See color plate section.


(ii) Second-degree injury is similar to axonotmesis above. The injury does

not disrupt the basal lamina or endoneurial integrity but the axon is

damaged and there is distal Wallerian-like degeneration. Recovery occurs

through regeneration of axons.

(iii) Third-degree injury involves a loss of axon continuity and subsequent

Wallerian-like degeneration but the perineurial sheath remains intact.

The general arrangement of the fascicles is preserved but there is

endoneurial disruption.

(iv) Fourth-degree injury involves a severe disruption of the peripheral nerve

trunk with disorganization of its topography and loss of fascicular

arrangement. The nerve trunk is in continuity because the epineurium

is not separated.

(v) Fifth-degree injury involves a complete transection or separation of the

entire nerve trunk, as described by neurotmesis above.

Mixed types of injuries and partial nerve injuries are common. In human

peripheral nerve injuries, a combination of local demyelination and some axonal

degeneration is usually the case. The prognosis for recovery depends on the

relative proportions of each and they can be estimated with electrophysiological

testing. For example, identifying conduction block across an injury segment,

with retained excitability of the distal nerve segment identifies local demyelina-

tion, and a good prognosis for recovery commensurate with remyelination.

If, however, loss of excitability of the distal nerve segment is identified, this

indicates Wallerian-like degeneration beyond the injury site and recovery

requires much slower axonal regeneration. Similarly, when motor axons undergo

degeneration distal to an injury site, the muscle fibers they innervate develop

abnormal spontaneous activity. This can be recorded by an electromyography

needle electrode: spontaneous fibrillation potentials (abnormal spontaneous

action potentials of single denervated muscle fibers) and related potentials

termed positive sharp waves. Loss of membrane excitability of the axons in the

distal nerve stump and the appearance of fibrillations in the denervated muscle,

however, require time to emerge; longer if the lesion is more proximal.

Overall then, the loss of excitability and Wallerian-like degeneration following

nerve trunk injury requires a suitable interval to begin (up to 10 days). The relative

proportion of damaged axons can then be judged by the amplitude of the com-

pound muscle or nerve action potential (CMAP) recorded distal to the injury

site in reference to normal control values or to an intact contralateral nerve

(see Chapter 4). Fibrillation potentials are recorded from muscle fibers that have

lost their connecting axons. In contrast, distal CMAPs are normal in a demyelinat-

ing injury, but there may be complete block of conduction across the injury

43Overall considerations

site (normal CMAP distal to injury, absent or attenuated CMAP proximal to the

injury). No axonal degeneration occurs and no fibrillations appear in the muscle

despite paralysis. A detailed description of electrophysiological approaches

appears in Chapter 4.

Waller, Wallerian degeneration and Wallerian-like degeneration

Augustus Waller [715] examined the sequence of pathological changes

that ensue in peripheral nerves distal to a transection. This complex process is

crucial for subsequent regeneration of axons. Waller writes:

During the four first days, after section of the hypoglossal nerve, no

change is observed in its structure. On the fifth day the tubes appear

more varicose than usual and the medulla more irregular. About the

tenth day the medulla forms disorganized, fusiform masses at intervals

and where the white substance of Schwann cannot be detected. These

alterations, which are most evident in the single tubules, may be found

also in the branches. After twelve or fifteen days many of the single

tubules have ceased to be visible, their granular medulla having been

removed by absorption.

Without the benefit of staining techniques that were later developed by

Golgi, Cajal, and others, Waller described the “fusiform” change of distal nerve

segments degenerating into what are now called “myelin ovoids.” These ovoids

represent globules of degenerative myelin and axon debris coursing along

the trajectory of the previously intact axon (Figure 3.4). Strictly speaking, and in

keeping with Waller’s original description, “Wallerian” degeneration refers to a

series of events in axons distal to a sharp nerve trunk transection (Figure 3.5). Local

damage of axons from other causes, including axonotmesis or neuropathies,

however, releases a nearly identical sequence of degenerative changes as those

described by Waller. The similarities have prompted many pathologists and others

to refer to all forms of axonal degeneration as “Wallerian degeneration.” We will

term other forms here as “Wallerian-like” degeneration. There are also important

similarities between these processes and those involved in pruning of redundant

regenerating axon branches or aberrant branches during development [430].

After an irreversible axon injury (presumably there are reversible types,

perhaps from transient depolarization after a blunt injury discussed above),

the axon segment distal to the injury site is rapidly signaled to begin degeneration.

This is an important concept because formerly, axonal degeneration was thought

to ensue from a simple cut off of “nutrients” supporting the distal axon segment.

Instead, a specific signal triggers breakdown. The exact signal is unknown but

it may involve a calcium transient.


Figure 3.4 An image of teased myelinated fibers illustrating Wallerian-like

degeneration (top) with breakdown of the axon and myelin and early formation

of digestion chambers. Wallerian-like degeneration occurs after axonotmesis injuries

and Wallerian degeneration after neurotmesis. The degenerating fiber sits above

an adjacent normal myelinated fiber.

Figure 3.5 An illustration of the sequence, from top to bottom, of Wallerian

degeneration and subsequent regenerative sprouting of a transected myelinated axon.

Larger rounded hematogenous macrophages invade the injury site in the fourth figure

from the top. The illustration is reproduced from the classical monograph by Tinel [681].

45Waller, Wallerian degeneration and Wallerian-like degeneration

Axonal breakdown also involves more persistent influx of calcium. The pro-

cess of breakdown is then initiated by E3/E4 ubiquitin ligases and proteases

including calpain (Figure 3.6). Microtubular dissolution is thought to occur first

[728,771], with consequent severe disruption of axonal transport and further

cycles of degeneration. Digestion of neurofilament lattices is usually completed

over the first 7–10 days and neurofilament material can be recognized immuno-

histochemically as disorganized fragments that eventually disappear. If examined

by semithin LM, the distal nerve stump may, in fact, look normal for a few days

depending on the type of nerve injury; intact and uncollapsed myelinated

profiles can be seen. Ultrastructural studies resolve this early phase of Wallerian

degeneration by identifying loss of normal axoplasm ultrastructure, then atro-

phy and disappearance of the axon. The sequence of axon degeneration appears

to occur first just beyond the site of axonal injury, or proximodistal (centrifugal

degeneration) but also at the distal terminals of the injured axons, or distoproximal

(centripetal degeneration) [235]. There are also retrograde changes proximal or above

Wld s = UFD2/E4 + Nmnat

Fusion protein

NAD

E3/E4 ubiquitin ligase + Calpain+ Calcium tr ansient

Disrupted microtubules

Neurofilament dissolution

Axon breakdown(+ m yelin breakdown if myelinated)Loss of excitability

PhagocytosisSC proliferation

Blunt orsharpinjury

SIRTIAxon

protection

1

2

Figure 3.6 A scheme illustrating some of the molecular events thought to

occur in active Wallerian degeneration. See the text for details. (Illustration by

Scott Rogers.)


the level of the injury itself known as “traumatic degeneration.” This can extend to the

first or second node of Ranvier of a myelinated fiber of the proximal stump.

In more severe instances, traumatic degeneration refers to retrograde loss of

the entire neuron, a topic discussed in Chapter 5 [195,572].

Despite the rapid sequence of early events in degenerating axons, it is inter-

esting that they retain their membrane excitability for a period of time after

nerve section, mentioned above. If a distal segment of a transected nerve is

stimulated, distal nerve action potentials or motor action potentials in motor

nerves can be recorded for several days. Excitability gradually declines. In the

clinical literature, a rule of thumb is that electrophysiological recordings of

injured nerves should be delayed for at least 7–10 days to glean the full extent

of their damage. By this time, all permanently damaged axons will have lost

excitability and an accurate appraisal can be made.

An early and active role for the ubiquitin–proteasome system (UPS) has also

been discovered during Wallerian degeneration. For example, inhibitors of the

UPS can delay it [164]. This important finding emerged following identification

of a spontaneous mouse mutant known as WldS with axons that survive and

persist for long periods after they have been transected [433]. In this mouse, the

genetic defect involves a dominant triplication overexpression of a ubiquitin

regulatory enzyme-related protein known as WldS. WldS constitutes a fusion

protein involving a 70-amino acid portion of UFD2/E4 (a protein involved in

protein ubiquination; the 70 amino acid portion, however, does not influence

ubiquination) and nicotinamide mononucleotide adenylyltransferase (Nmnat,

an enzyme involved in nicotinamide metabolism [430]). Elevated Nmnat activity

from overexpression of the fusion protein is then thought to thereby increase

levels of nicotinamide adenine dinucleotide (NAD). NAD, in turn, acts as an

axonal protectant through a protein deacetylase known as SIRT1 (a member of

the Sir family of protein deacetylases and mammalian ortholog of Sir2) [21,598].

All of these players thus orchestrate a balance between retention of axonal

integrity and active axonal dissolution (Figure 3.6). The administration of NAD

(replicating the impact of overexpresion of WldS) or enhancement of Sir2 activity

though a specific pharmacological agent, resveratrol, was shown to slow axonal

degeneration after transection. Exactly how these interventions act to protect

axonal integrity, however, is not completely established.

Wallerian and Wallerian-like degeneration share characteristic morpho-

logical features that can be recognized (and counted) in high-quality histological

samples (Figures 3.4, 3.7). The techniques used to demonstrate the features

include the use of teased nerve axon preparations stained with osmium tetroxide,

or semithin transverse or longitudinal sections of nerve that are plastic, or epon

embedded (see Chapter 4). Typically, the first change is the disappearance of the


fine structure of the axons (neurofilaments, microtubules, see above). Axons

become watery in appearance, then swell, or disappear within the confines of

the myelin sheath. In transverse section, what might appear to be “normal”

myelinated profiles at low power magnification are actually profiles with

shrunken or nonexistent axoplasm when examined by EM. Within each inter-

node, axon and myelin contents retract into bulb-like “digestion chambers”

Figure 3.7 Transverse sections of peripheral nerves showing myelinated axon

profiles undergoing Wallerian-like degeneration. The white arrows identify

degenerating profiles earlier (top) and later (bottom) following injury. In the lower

panel, large macrophage profiles containing myelin and axon breakdown products

are found amidst newly regenerating axons (arrowheads). The images are from

a semithin sections, stained with toluidine blue, from rat sciatic and sural nerves.


or oblong structures consisting of myelin and axon debris. These digestion

chambers are called “E” fibers in the Dyck classification of teased single myeli-

nated axons [159]. Digestion chambers appear along the full length of the severed

or otherwise damaged nerve. They likely represent “phagocytosis” by nearby

Schwann cells, although later involvement by both resident and invading macro-

phages play a role.

Morris and colleagues [477] explored the early role of SCs in breakdown of

myelin at the ultrastructural level during retrograde degeneration in proximal

nerve stumps after section, a process closely related to Wallerian degeneration.

The authors suggested that SCs are responsible for digesting their own myelin

after injury and sometimes assume a configuration described as “transformed

Schwann cells.” Early changes included wrinkling and distortion of myelin

lamellae, erosion of circumferential lamellae, and myelin fragmentation.

Myelin retracted from the nodes and myelin debris was incorporated into vacu-

oles with membranous material and pale globular inclusions. Hematogenous

macrophages invaded nerve injury zones but only within restricted time frames

(5–7 days following injury) and only transiently penetrated into sites in the distal

degenerating nerve. This process has subsequently been beautifully illustrated

using an iron labeling MR technique [38].

In transverse sections, axons and their associated SCs undergoing degener-

ation are recognized as swollen irregular structures containing myelin debris,

i.e., digestion chambers described above. Such profiles may be identified as

single degenerating axons, within sectors of the nerve fascicle or may be perva-

sive, involving all axons of the nerve trunk (e.g., after neurotmesis). At early

stages during the removal of myelin debris, SCs and macrophages can be identi-

fied heavily laden with cholesterol vacuoles. With time, digestion chambers

containing axon and myelin debris atrophy. If longitudinal sections or teased

fibers are examined at these later time points, such digestion chambers appear

along the length of the original fiber like beads on a string linked by collagen

fibrils. These late “E” fibers may persist for years in the nerves of some patients

with chronic neuropathies.

The progress of Wallerian degeneration critically depends on the behavior

of SCs. SCs alter their relationship with the stable myelin sheaths they maintain

and instead initate myelin breakdown. Neuregulins (NRGs) are described in

more detail in Chapter 5, but they act as local signaling messengers released

from axons during nerve injury and regeneration. In a series of events linked

to NRG signaling (discussed further in Chapter 5) through its erbB2 SC receptor,

SCs change their phenotype. For example, they undergo dedifferentiation and

downregulation of myelin protein synthesis within 2 days of the loss of axon

contact [380]. Constitutively, erbB2 is expressed at nodes of Ranvier flanked


by the proteins Ezrin and Caspr and it is thought to signal from the SC microvilli

that are in contact with axons. Interestingly, activated erbB2 is also expressed

in SCs associated with unmyelinated axons with or without injury. ErbB2 is

activated through phosphorylation within minutes of an axonal injury then

declines only to again be re-expressed several days later, a time frame that

coincides with SC proliferation [242]. ErbB2 activation occurs through the inter-

mediate early gene c-jun. In summary, SCs initiate the self-destruction of their

myelin sheaths, followed by changes in behavior that include later migration

and proliferation.

Wallerian degeneration also depends on the expression of pro-inflammatory

cytokines and chemokines (see review [647]). Expression of anti-inflammatory

cytokines can occur concurrently. Macrophage entry is a critical event in

the process. After sciatic nerve crush, T lymphocytes and macrophages infiltrate

the lesion site within 48h and into the distal nerve stump by day 4

[69,303,304,552,645,646]. There is a biphasic pattern of rises in the expression

of several chemokines and cytokines distal to a sciatic nerve transection at day 1

and again at day 14 [551]. The pro-inflammatory chemokines MCP-1 (monocyte

chemoattractant protein-1) and MIP-1a (macrophage inflammatory protein 1a)

have prominent early rises, especially in the case of MCP-1 just adjacent to the

injury site. Anti-inflammatory cytokines, TGF-b1 and IL-10 have similar changes.

IL-1b had two types of increase. An early rise occurred close to the injury site,

while a later rise developed in segments further from injury. At the mRNA level,

similar patterns of expression have been identified [647]: rises in IL-1b, IL-6, and

IL-10 peaking 1 day after crush with a more sustained later upregulation of IL-18,

IFN-g, TNF-a, and IL-12p40 out to 14 days. TNF-a, a central player, is expressed

in Schwann cells, fibroblasts, endothelial cells, and macrophages [648,713].

Wallerian degeneration is also associated with upregulation of matrix metallo-

proteinases (MMPs) that act to degrade inhibitor constituents of the extracellular

matrix, particularly chondroitin sulphate proteoglycan (CSPG). In peripheral

nerve, MMP-2 and MMP-9 rise during Wallerian degeneration and, with laminin,

facilitate the later ingrowth of axons [180] (see Chapter 10).

The particular and coordinated pattern of inflammatory molecule expression

appears to be intimately related to the success of Wallerian degeneration and

the myelin clearance that is part of it. For example, the continuous infusion

of neutralizing antibodies to either the combination of MCP-1 and MIP-1a or to

IL-1b within the microenvironment of the nerve undergoing Wallerian degener-

ation was associated with reduced numbers of phagocytic macrophages and

preservation of myelin sheaths (but not necessarily axons). Similarly, mice

lacking IL-6 had delayed sensory axon regeneration [778]. Mice lacking TNF-a

had reduced macrophage influx [404].


Specific macrophage chemoattractants also include MCP-1 discussed above,

LIF (leukemia inhibitory factor), and pancreatitis-associated protein III (PAP-III)

released from Schwann cells [497,655,682]. Rats with knockdown of PAP-III

had slowed macrophage recruitment into injured peripheral nerves and a delay

in subsequent regeneration. SCs may be activated by IL-6 in turn synthesizing

both LIF and MCP-1 that serve as macrophage attractants. This interesting path-

way also involves an autocrine loop whereby LIF induces MCP-1 expression

as well [682].

Several additional molecules have important roles in mediating axon break-

down. Nitric oxide, released by rises in the expression of inducible nitric oxide

synthase (iNOS) in Schwann cells and macrophages at the site of a peripheral

nerve injury [391,799] is one such mediator. Nitric oxide reacts with the super-

oxide radical to form peroxynitrite, a species capable of lipid myelin peroxida-

tion, an important early step in its eventual clearance during Wallerian

degeneration [600,704]. Mice lacking iNOS have delays in the progression of

Wallerian-like degeneration and in subsequent reinnervation [392]. Interest-

ingly, erythropoietin (see Chapter 9), generated by SCs, in response to ambient

NO production after injury is paradoxically capable of protecting axons and

preventing axonal degeneration [338]. NO signals thus have varying roles

depending on their timing, intensity, and targets.

RAGE (receptor for advanced glycosylation endproducts) mediates mononuclear

phagocyte participation in Wallerian degeneration. Phagocytes lacking proper

RAGE signaling had decreased p44/p42 MAP kinase phosphorylation and mice

with a dominant negative RAGE had delays in the progression of Wallerian

degeneration and in subsequent regeneration [584].

Fewer proteins have been identified that normally dampen Wallerian

degeneration. Desert hedgehog protein (Dhh) may be one example: mice lacking

it appeared to have accelerated Wallerian degeneration. It is unclear how

this signaling pathway integrates with others to regulate the overall rate of

degeneration [29].

Neuromas

Neuromas are swollen distorted portions of a nerve trunk that result

from severe traumatic disruption. Examples are given in Figure 3.8. “End nerve”

neuromas are found at the proximal stump end of a completely interrupted

peripheral nerve (neurotmesis). At one time, neuromas were thought to repre-

sent enlarged masses of aberrantly sprouting axon profiles misdirected to grow

back on themselves. While misdirected axons are a common feature of neur-

omas, it should be emphasized that they are not simply composed of dense

51Neuromas

axon bundles. Neuromas are complex structures (Figures 3.9, 3.10). They

contain former fascicles, infiltrating macrophages, abundant connective tissue

with dense collagen deposition, and avascular zones likely resulting from

local ischemia [755,787,807,809]. Angiogenesis also occurs within neuromas,

and accompanies a pancellular “burst” of cellular proliferation that develops

in their first 5–7 days [805]. Other proliferating cells include mast cells, macro-

phages, SCs, and fibroblasts. Thus, beyond simple ingrowth, the complex

changes that occur in neuromas may involve ischemia, protease activity, local

free radical release, and elaboration of growth factors from macrophages and

mast cells. They also likely include signals from the extracellular matrix. All

of these constituents may play roles in generating neuropathic pain from

neuromas.

At early stages, neuromas exhibit other interesting features that influence

their structure and behavior. Transected axons within them form swollen

endbulbs or boutons, also discussed in Chapter 5, presumably from accumulated

products of anterograde transport and impaired turnaround transport. These

endbulbs accumulate biologically active molecules that include neuropeptides,

opioid receptors, nitric oxide synthases, sodium channels, and others.

Neuromas “in continuity” arise from partially injured nerves (axonotmesis) in

which one or several fascicles do not regenerate (Figure 3.8). Instead, the axons

Figure 3.8 Illustrations of several types of chronic nerve trunk injuries from the

classical monograph by Tinel. On the left (I) is a neurotmesis injury with an end

neuroma on the proximal (top) stump. To the right are examples of progressively

more severe axonotmesis injuries with neuromas in continuity. It is possible that

injures 3 and 4 have a small number of regenerating axons traversing the injury

site [681].


and connective tissue of these fascicles form a neuroma as a local enlargement

along the nerve trunk. Such lesions provide an important lesson in regeneration

neurobiology. This is a scenario in which a nerve trunk remains connected, yet

axons interrupted by the injury fail subsequently to grow into the distal portion

of the nerve. There is ample guidance by connective tissue to indicate where

axons and Schwann cells should travel, yet inexplicably, both form misoriented

aberrant neuromas. The lesson is that anatomical reconnection or guidance

from connective tissues is insufficient for successful regeneration. The milieu

must provide additional signals and guidance to support axonal growth beyond

simple continuity.

Figure 3.9 Longitudinal section of a neuroma in continuity as illustrated by Tinel.

Note the complex disorganization of the neuroma and that most axons from the

proximal stump (top) do not reach into the distal stump [681].

53Neuromas

Partial forms of peripheral nerve injury

Chronic constriction injury

There are several partial forms of peripheral nerve trunk injury.

The chronic constriction injury (CCI), first described by Bennett and Xie [41],

has been widely studied as a model of neuropathic pain. The lesion involves the

placement of four loose ligatures around the trunk of the sciatic nerve in rats or

mice. The local inflammatory reaction from chromic catgut (dissolvable) sutures

causes swelling of the nerve trunk within 2–3 days. The segments of the nerve

between the ligatures become ischemic and induce axonal degeneration in a

large proportion of the axons. CCI develops more gradually than simple crush or

transection. Most likely, an interaction between damaged axons, en passant intact

Figure 3.10 Details of the neuroma in continuity from 3.9 by Tinel. Note the

presence of minifascicles and the absence of a consistent trajectory direction for

axons within this structure [681].


neighbors, and the release of a barrage of inflammatory mediators is responsible

for generating neuropathic pain. Once events in the nerve trunk initiate pain,

changes in neuron phenotype, their upstream pathways and glia of the ganglion,

dorsal horn of the spinal cord and higher then follow [696]. Large myelinated

axons are the most susceptible to CCI whereas a proportion of small unmyeli-

nated axons are not injured. Spared smaller caliber axons may allow selective

sensory information to traverse the CCI lesion and in turn generate the pain

syndrome associated with it. Alternatively, some of the pain behavior may arise

from ectopic discharges of injured axons, rather than spared ones. Ectopic or

spontaneous activity of large caliber axons, for example, can be detected after

CCI [393]. If tested for, behavioral features of neuropathic pain emerge by 48h

and persist over 3 weeks.

Standard measures used to identify pain include the latency of paw with-

drawal on the side of the nerve injury to a thermal stimulus. More rapid

withdrawal indicates hyperalgesia, whereas delayed withdrawal indicates sens-

ory loss. Allodynia refers to the perception of a normally innocuous stimulus as

painful. Mechanical allodynia is tested by determining the degree of withdrawal

to a mechanical filament probe of graded bendable force. A greater response

than expected to smaller caliber (more easily bent) filaments indicates allodynia

(see Chapter 4). At and after 3 weeks following a CCI injury, there is evidence that

axon regeneration has begun. Some axons grow around the area of the sutures

by penetrating through the epineurium and growing on the outside of the nerve

trunk to reach the distal stump. The behavioral measures associated with pain

can also be used to track regeneration of sensory axons. Pain behavior dimini-

shes as regeneration proceeds, a time when the local exposure of axons to

inflammatory mediators declines.

CCI has been an important model of pain neurobiology and has been widely

used to address mechanisms or pharmacological approaches toward treatment.

The lesion is partial, probably resembling common types of human injuries.

For example, severe neuropathies that occur following trauma in some instances

may involve similar types of localized ischemia at the site of injury. Compression

may occur from a bone fragment (e.g., radial nerve injuries from humeral

fracture), from a hematoma (e.g., sciatic injury from a pelvic fracture), or from

localized ischemia during a compartment syndrome.

A related partial nerve injury pain model resembles the CCI model and

involves placement of a single tight (in contrast to the loose ligatures of CCI)

ligature partway through a portion of the nerve trunk. The model generates

crush and inflammation of one portion of the nerve trunk adjacent to an intact

and uninjured portion. Finally, the Chung model of neuropathic pain involves

the placement of a tight ligature around a single nerve root, L5, supplying the

55Partial forms of peripheral nerve injury

distal sciatic nerve [115,342]. In this model the intact and injured nerve root

axons combine through the nerve plexus as the fibers enter the mixed sciatic

nerve trunk. Downstream in the sciatic nerve, therefore, there is an admixture of

intact axons from uninjured nerve roots and axons undergoing degeneration

from L5. En passant interactions of intact axons with degenerating neighbor

neurons and inflammatory byproducts of Wallerian-like degeneration may

generate ectopic axon discharges associated with pain behavior.

Compression

Compression injuries are common in patients. Forms of chronic com-

pression are known as “entrapment” neuropathies. Sites of frequent involve-

ment in humans include the carpal tunnel at the wrist for the median nerve,

the cubital tunnel for the ulnar nerve at the elbow or the exposed fibular head

at the knee for the peroneal nerve. Less commonly, a bone fragment, hematoma,

or tumor may acutely compress a nerve. There has been a tacit assumption that

such lesions are ischemic since they may interrupt the arterial supply or prevent

venous drainage. As will be discussed in Chapter 7, however, there is very little

actual evidence for significant ischemia in most of these common entrapment

neuropathies. Since the vascular plexus of the peripheral nerve trunk is highly

redundant from multiple feeding vessels and provides flow in excess of meta-

bolic needs, it is difficult to generate local ischemia from a single pressure point.

Stretch

Stretch injuries are the least studied form of partial nerve injury. The

robust tensile properties of longitudinally oriented collagen fibers protect axons

from stretch. In addition, axons have redundant curved trajectories within the

nerve trunk, allowing them to lengthen substantially before they can be pulled

apart. Stretch injuries, unlike those above, may involve a combination of ische-

mia from disruption of the epineurial plexus, and direct mechanical distortion

of axons. They are some of the most common human nerve injuries.

Direct injuries to sensory ganglia

Direct injuries of the dorsal root ganglia (DRG), housing sensory

neurons, have not been characterized in the clinical literature. Their importance

and prevalence are unknown. It is possible, however, that in the cervical and

lumbar regions of humans chronic compression or traumatic injury of DRG

can generate forms of spinal or radicular pain [12,410,481]. DRGs are not gener-

ally sampled in clinical studies. DRGs, however, may be more vulnerable to

ischemia than nerve trunks. Ischemic lesions of the ganglia result in down-

stream Wallerian-like degeneration of their distal axon branches [753]. Lesions


of peripheral branches of sensory dorsal root ganglia neurons are associated

with retrograde cell body reactions and changes in phenotype, discussed in

Chapter 5 [403].

Blood nerve and ganglion barriers and injury

Wallerian-like degeneration is associated with loss of integrity of the

perineurial component of the blood nerve barrier [731,732]. This may be import-

ant to allow inflammatory cells and blood borne trophic molecules access to the

injury site and to permit better clearance of axon and myelin debris. The change

also has important implications for future therapeutic interventions where

penetration into injured and regenerating fascicles is critical.

Summary

Peripheral nerve injuries can vary tremendously in their severity and in

their likelihood to recover. While several classifications of nerve injury have

been described, the most important consideration is whether myelin alone is

disrupted or whether there is additional axon damage. Injuries that only have

demyelination, known as neurapraxia, have conduction block across the injury

site but can recover relatively rapidly. In contrast, axon damage from either

blunt or penetrating damage leads to a series of events known as Wallerian or

Wallerian-like degeneration. Subsequent recovery requires axon regrowth that

involves a much slower pace of recovery. The Wallerian degenerative process is

not passive but an active pathway in which molecular players are only now being

identified. Axons undergo a series of structural changes with loss of excitability

and eventual phagocytosis while local chemokines, cytokines, and molecules

expressed in axon endbulbs have an important influence. The characteristics

of injury therefore have a substantial impact on the local milieu of an injured

nerve trunk and provide a context for all future regenerative activity, considered

in the next few chapters.

Suggested reading

Griffin, J. W., George, E. G., Hsieh, S. T., & Glass, J. D. (1995). Axonal degeneration and

disorders of the axonal cytoskeleton. In The Axon. Structure, Function and

Pathophysiology, Chapter 20. Oxford, UK: Oxford University Press [235].

Luo, L, O’Leary, D. D. M. (2005). Axon retraction and degeneration in development and

disease. Annual Review of Neuroscience, 28, 127–156 [430].

Sunderland, S. (1978). Nerves and Nerve Injuries. 2nd edn. Edinburgh, UK: Churchill

Livingstone [660].

57Summary

4

Addressing nerve regeneration

The pace of molecular discovery relevant to nerve regeneration has

accelerated. New insights into regeneration, however, have not necessarily been

partnered with rigorous approaches to measure regeneration. The purpose of

this chapter is to engender readers with a healthy appreciation of new findings,

based on rigorous approaches, that confirm the complexity and beauty of the

regenerative process. Similarly, the reader should be skeptical of approaches that

do not live up to that standard. Assays of regeneration should ideally encompass

all of the crucial steps involved in the regenerative timetable: early sprouting,

axon elongation, regrowth of axon radial caliber or girth, remyelination of larger

caliber axons, repopulation of nerve trunks by mature axons, and extension to

target tissues. During regeneration axons regain electrophysiological properties

that they have lost, features that can be carefully assayed. Finally, it is critical to

know whether there has been a resumption of function, addressed through

“functional” or behavioral endpoints. This chapter presents a summary of regener-

ation assays and a discussion of their strengths and limitations.

Structural (histological) approaches

Few other measures can convey the structural beauty of regeneration

captured in a high-quality histological snapshot. Histological techniques

demand strict attention toward the details of specific protocols and they require

optimal handling of specimens that are appropriately sampled. Their exactitude

sets a standard of quality that is enormously satisfying. Unfortunately, classical

histological approaches are frequently dismissed and substituted with easier

or more colorful techniques that have lower resolution. Classical histology

continues to be used in clinical laboratories to address disease processes.

58

In regeneration, use of semithin plastic embedded sections sampled at a fixed

distance from an injury site offers an assessment of the numbers and maturity of

newly regenerating axons, both myelinated and unmyelinated. They also address

the overall architecture of the nerve and can be used to identify mast cells,

macrophages, and axons undergoing Wallerian-like degeneration.

I will begin with some very practical issues. The benefits of careful handling and

fixation cannot be overemphasized. Nerves must be gently manipulated and not

pulled or stretched! If specific portions of a nerve are to be harvested or biopsied,

forceps should be used only on the very ends of the specimen. Sharp scalpels

or microscissors must be used for severing the nerves. Rough handling can be

recognized by microscopists. For example, sectors of fascicles have excessively

folded and crinkled myelin sheaths, often obscuring the axon within. Some

portions of the axon may appear paradoxically denuded of myelin. This appear-

ance results from telescoping of compressedmyelinated fibers during harvesting.

In some cases, further sections taken deeper into the epon block may save the

specimen because they are beyond and bypass the damaged area.

Specimens must be kept moist immediately after resection and before they

are fixed. Nerves, particularly those of mice, dry very quickly. The sheaths of

myelinated axons swell, lose their contrast and with time completely distort the

myelinated axon profile [438] (Figure 4.1). These nerves contain numerous lightly

stained donut-like concentric circles without any visible axons visible within

them. Our approach to prevent drying, otherwise known as desiccation artifact, is

to rapidly place specimens in iso-osmolar normal saline and then fixate quickly.

There are a number of classical histological approaches used to study peripheral

nerves. Whilemany are appropriate for clinical biopsy laboratories, they are not all

helpful for studying regeneration. These methods include paraffin embedding of

thick sections and staining with hematoxylin and eosin, myelin stains (Luxol fast

blue and others), or silver axon stains such as Bielschowsky or Bodian. The major

drawback of paraffin-embedded material is the inability to resolve individual

myelinated axons reliably, especially small ones. Silver-stained axon profiles may

be confused with collagen profiles. Accurate measures of caliber are not possible.

None of the paraffin methods offers the resolution of glutaraldehyde fixed and

osmium-stained epon or plastic-embedded semithin sections that we clearly favor.

With this acknowledged bias, there are variations in how semithin sections

can be made, all with excellent results if carried out meticulously. I will relay

the approach of Dyck and colleagues that is used for both clinical and research

purposes. The detailed histological protocol is provided in Appendix 4a.

It involves fixation in cacodylate-buffered glutaraldehyde fixation, followed by

alcohol dehydration, osmium tetroxide staining, and finally epon (plastic)

embedding. An ultramicrotome provides LM (light microscopy) semithin sections

59Structural (histological) approaches

of 0.5 to 1.0 micron diameter that outline the full structure of the nerve trunk in

exquisite detail, including internal axonal structure. Transverse LM sections

provide measures of myelinated axon (MF) numbers, axon caliber (area, diameter),

and myelin thickness. From these measures, distributions of axon size can be

plotted as histograms discussed in Chapter 2. Information about myelin thick-

ness can also be given using the G ratio, the axon diameter divided by the

diameter of the whole fiber including myelin sheath (see Chapter 2).

The same epon-embedded samples used for LM can be used to provide ultra-

microtome thin sections that are laid on grids for electron microscopy (EM).

EMs, in turn, are used to measure the density and caliber of unmyelinated axons or

fibers (UFs). As with MFs, construction of fiber size histograms of unmyelinated

axons generate additional information. Both LM and EM are used to examine

blood vessel numbers and structure, basement membrane thickness, macrophage

or mast cell numbers, and other features.

(a)

(c)

(b)

(d)

Figure 4.1 Transverse sections of normal sciatic nerves from mice illustrating

desiccation (drying) artifact. A is from a properly preserved and fixed nerve. In B, note

that the individual axons within myelinated profiles have shrunk and retracted.

In C and D, severe desiccation artifact completely distorts the profiles andmakes these

nerves unsuitable for further analysis. (Reproduced with permission from [438].)

60 Addressing nerve regeneration

Dyck and colleagues also recognized the importance of eliminating artifacts

from fixation of nerve [527–530]. The use of hyperosmolar fixation in pathology

laboratories that emphasize CNS work produces a very common artifact.

The fixative shrinks and distorts the axon profiles generating images of highly

irregular (but artifactual) myelin sheaths that are collapsed into the axon.

The substitution of an iso-osmolar fixation matches the osmolality of the periph-

eral nerve to that of the fixative and preserves features of the nerve architecture.

Despite these long-established protocols, many examples of poorly processed

nerves continue to be published.

There are three basic ways to fix nerves: perfusion fixation, in situ fixation,

and immersion fixation. Perfusion fixation involves the placement of a cannula

in the heart (or other great vessel, protocols vary) of an anaesthetized animal and

an infusion pump delivers titrated fixative appropriate for the animal size. Blood

and fixative are drained from the femoral vein. Many laboratories combine this

method with immersion fixation because perfusion may not adequately fix all

the tissues. Immersion fixation simply involves placing the specimen (nerves can

be loosely tied with nylon to small sticks and a suture placed on the distal end

for orientation) in a vial of freshly prepared fixative. Perfusion fixation makes

detailed dissection of ganglia and nerves more difficult since the tissue is

hardened, discolored, and without blood. Immersion fixation alone may not

properly fix larger tissues like human ganglia or rodent spinal cord. This limita-

tion may be circumvented by first dividing the tissues into smaller portions and

prolonging the fixation period. In situ fixation offers a compromise: the nerve,

cord or ganglion is exposed and the bed has fixative applied and left for a period

of time (usually 30 minutes). After this fixation is completed, the tissues may be

removed and further fixed by immersion [603].

Image analysis programs are available including free software for PCs

originally developed by NIH (Scion Image, Scion Corporation, Frederick MD;

www.scioncorp.com and Image J, http://rsb.info.nih.gov/ij/). Rigorous calibration

for magnification is a requirement of all software approaches. Next for consider-

ation is the extent of the regenerating nerve that should be examined. Some

reports have emphasized measurements of the density of remyelinating axons

measured distal to an injury. In this approach, selected (arbitrary or random)

fields undergo measurement and the behavior of the whole specimen is inferred.

Density measurements require less time to complete and consequently are

easier. Although density and measurements of total nerve axon numbers may

correlate directly with one another, this is not a fixed rule. Indeed, some regener-

ating nerves may have fields with high densities of regenerating axons but

overall have not succeeded in regenerating as many axons. Regenerating nerves

sometimes exhibit clusters or minifascicles that demonstrate this disparity.

61Structural (histological) approaches

Small minifascicles may represent sprouts from single parent axons. Thus, a

density measure might spuriously suggest high fiber outgrowth but an exami-

nation of the entire nerve field would indicate otherwise. In smaller nerve

trunks or during early regeneration, measuring the total number of regenerated

myelinated axons might provide the most rigorous appraisal of the axon popu-

lation. In other instances, measuring the density of myelinated axons when large

numbers of axons are present may be more feasible. Finally, fiber size histograms

can be used to sort out whether there is a subpopulation of axons with interest-

ing behavior. These subpopulations might not be captured or understood by

relying only on overall measures of mean axon caliber.

Our approach is to select early time points measuring the numbers of all new

axons distal to the injury site and to emphasize distal branches of the sciatic

nerve. In this way, the technician might not be overwhelmed by counting large

numbers of axons. In the case of unmyelinated axons presented by EM, it is

daunting to count entire nerve trunks or fascicles. Density measures in random-

ized sampling fields are required. The quality of peripheral nerve EMs can

be judged by whether individual myelin lamellae of myelinated axons can be

resolved and distinguished under higher power.

Transverse sections of the nerve at a fixed distance distal from the injury site

(e.g., 5–15mm in a rat sciatic or sural nerve) allow for the estimation of axon

outgrowth after injury. As alluded to above, axons identified in this way may

arise from a smaller complement of parent axons, i.e., they are multiple regenera-

tive sprouts. These sprouts may be myelinated or unmyelinated axons. Therefore,

correlating these kinds of measures with another approach, such as fluorchrome

backlabeling (see below) may help to determine the number of actual parent

neurons generating new axons.

During early regeneration, myelinated axons are smaller in caliber and

have thinner myelin sheaths than mature fibers (higher G ratio). In very early

regeneration, none of the axons may have begun to myelinate. The number of

regenerative clusters (groups of myelinated axons in minifascicles) may also be

measured. Finally, it may be of interest to measure the number of axons under-

going Wallerian-like degeneration. In some conditions, delayed Wallerian-like

degeneration can be identified by abnormal persistence of degenerating profiles.

Regenerative success, in turn, requires their prompt and efficient clearance.

Immunohistochemical approaches

Immunohistochemistry is likely the most widely used assay of in vivo

axon regrowth. Its advantages are several. Immunohistochemistry allows label-

ing using specific molecular markers of axons, SCs, and other peripheral nerve


components. These may label subsets of axons, for example, Substance P (SP) for

sensory axons, choline acetyltransferase (ChAT) for motor axons. The caveat is

that many molecules are shared by both fiber types, e.g., CGRP. The disadvan-

tages of immunohistochemistry are that some antigens can be capricious to

label; they are nicely demonstrated with some fixatives and antibodies, yet

elusive to other approaches. Most immunohistochemistry uses immersion fix-

ation but fixatives vary and might include, for example, brief acetone fixation of

fresh samples, variations of paraformaldehyde fixation (e.g., Zamboni’s fixative),

and “antigen retrieval” protocols. These are not described here.

There are two approaches for using an immunohistochemical assay of nerve

regeneration. As in the LM approach, transverse sections of nerve at a fixed

distance from the injury site can be used to label numbers of regrowing axon

profiles. Since antigens are rarely evenly distributed through the diameter of an

individual axon, these profiles cannot be used to measure axon diameter or

caliber. Specific labeling of myelin proteins, such as MBP can be used to identify

myelinated fibers. The second approach is to study longitudinal sections of nerve

that may include the injury site. This approach, if used at early time points, can

be used to measure the number and outgrowth lengths of axons beyond the

injury zone.

There are several important considerations in using longitudinal sections of

outgrowing axons. Axons and their endoneurial fascicles bend and twist along

the nerve in three dimensions. Given this normal “wavy” trajectory, it may be

impossible, for example, to trace the origin of a given regenerating sprout to

its parent axon. Axons wind their way through larger distances than even thick

50-micron sections studied by confocal microscopy might sample. Thus, beyond a

few individual axons, a comprehensive ascertainment of branching in vivo is

often not feasible. Also, if only one–two high resolution thin sections are used,

it may be incorrect to infer the numbers of axons that are regrowing because

many axons may not have been captured in the section through the nerve;

several sections through the estimated center of the section are routinely

required. If the end of the section cuts through perineurium (because the nerve

trunk is curved or embedding is not flat), there may be a false sense that axon

growth has ended (i.e., false end of the fascicle). A marker with high resolution

(e.g., bIII tubulin; see below) may be associated with such dense labeling that

individual profiles cannot be counted easily. Overly dense labeling occurs when

analyzing the proximal stump of an injured nerve or the regenerating zone at

later time points. Finally, some axon antigens disappear only slowly distal to

injury during axonal degeneration. Residual degenerating material might then

be counted inaccurately as new axons. In most instances, however, these can be

distinguished as punctate nonlinear degenerative profiles.

63Immunohistochemical approaches

Some investigators have attempted to quantitate axon outgrowth bymeasuring

the labeled area from a whole nerve trunk specimen without analyzing individual

profiles, e.g., a measure of percent occupied area. While the approach appears

deceptively easy, it lacks satisfying resolution and can be distorted by a series

of artifacts. These artifacts are variation in sampled area, nonspecific labeling

by primary or secondary antibodies, overlap of multiple profiles, and others.

We think this approach yields inaccurate results and should be discouraged.

An approach that directs the proximal and distal stump of a transected

peripheral nerve (e.g., sciatic nerve of an adult rat) into a regenerative conduit

[452] permits the analysis of early nerve regenerative events (Figure 4.2). New

axons regenerate by growing along a connective tissue/fibrin bridge connecting

the proximal and distal stump. These can be resolved, counted, and measured

within the first few days following injury. Moreover, new axon outgrowth occurs

independent of the distal nerve stump and products of Wallerian degeneration.

At 7 days, for example, outgrowth is not dense enough to obscure individual

Figure 4.2 Diagram illustrating a setup to analyze early regenerating axons

across a peripheral nerve transection gap and to manipulate the regenerative

microenvironment with a subcutaneous access port. (Illustration by David

McDonald and reproduced with permission from [452].)


profiles: these can be counted at serial distances from the proximal stump to

the furthest distance of outgrowth. Both the numbers of outgrowing axons and

the longest distance traveled by an individual axon can thereby be estimated.

Some discussion here is warranted of what axon labels are most suitable for

addressing nerve regrowth. Neurofilament labels, such as the Nf200-antigen, the

heavy unit of neurofilament intermediate protein, are robust and address a large

proportion of axons [103,491]. Neurofilaments are heavily represented in larger

caliber axons, whereas they are present in much lower densities in small myeli-

nated and unmyelinated axons. Neurofilaments also do not extend to the distal

end of the axon and growth cone. Overall, a neurofilament label might therefore

underestimate the full extent of axon outgrowth. It may also not label small

caliber new axons with enough intensity to be detected. We have estimated that

Nf200 labels approximately 25% fewer distal outgrowing axons than a bIII

tubulin label. Despite this shortcoming, Nf200 labeling did not substantially

underestimate the distance the axons traveled. The distal nerve terminals not

labeled by Nf200 were highly variable in their trajectories, perhaps representing

unstable new sprouts sampling their microenvironment. Nf200 labels had larger

more discrete calibers more amenable to analysis. Overall, Nf200 might be

considered a more robust measure of “stable” axon profiles. While products of

neurofilament may persist for several days during Wallerian degeneration distal

to axon interruption, their profiles can be distinguished from that of viable axons.

By 7 days, for example, neurofilament profiles are short, broken, discrete, and

segmented in contrast to more continuous profiles of intact regenerating axons.

PGP 9.5 and bIII tubulin are two constituents that can be labeled in all

outgrowing axons. PGP 9.5 labels a ubiquitin hydrolase enzyme that is present

in all axons. It is the label of choice for addressing epidermal skin innervation

(see below). The bIII tubulin antigen is derived from subunits of microtubules.

Both molecules extend to the distal axon and growth cone. Their small distal

irregular profiles, however, may be difficult to evaluate as individual axons and

likely to represent unstable new branches. In longitudinal sections at, and distal

to, a crush zone, both markers require several days to disappear during phago-

cytosis and axonal degeneration. Thus, like neurofilament labels, PGP 9.5 and

bIII tubulin must be studied with care at early time points especially after crush.

Distinguishing residual material after crush from degenerating axons adjacent

to newly ingrowing axons can be challenging.

Several investigators have chosen GAP43/B50 as their label of choice for

examining outgrowing axons. Growth cones are preferentially decorated with

this growth-associated protein. Disadvantages are that more mature and stable

proximal profiles might not be labeled. A detailed comparison of its sensitivity

with neurofilament, PGP 9.5 and bIII tubulin has not been published. SCs also

65Immunohistochemical approaches

express the antigen and since they elaborate fine processes during regeneration,

they may be mistaken for axons. Other markers used to label specific axon sub-

populations include CGRP, galanin, IB-4, and other peptides [334,401,691,736].

A novel approach toward evaluating nerve regeneration involves the use of

YFP transgenic mice. The neuron specific thy-1 promoter is used to drive the

expression of YFP. While not all axons assume fluorescence, the majority do, and

the mice have provided elegant approaches toward tracing cutaneous nerve

behavior [102] or counting outgrowth of axons beyond injuries [170,744].

Electrophysiological measures of regeneration

There is a range of electrophysiological approaches used to measure

peripheral nerve regeneration in animal models or humans. One of the most

widely used is multifiber recording from motor and sensory axons. The method is also

described as a “population recording” from axons within a peripheral nerve

trunk. They are carried out under simple anesthesia in rodents or other mammals

and without sedation in humans. The recordings are based on the premise that a

supramaximal stimulation of a whole nerve trunk will recruit all electrically

excitable axons adjacent to the stimulating electrodes. Downstream of the stimu-

lating site, compound nerve action potentials (NAPs) are recorded that represent

summated individual axon action potentials. The amplitude (e.g., baseline to

peak) has an excellent correlation with the number of excitable axons that

have been recruited by the stimulation and that course beneath the site of the

recording electrode. Careful measurement of the distance between the stimulat-

ing cathode and the proximal pole (G1) of the recording electrodes provides

conduction velocity (CV) measurements. By convention, CV ¼ distance/latency

(time between stimulation and recording). Latencies are calculated between the

onset of the sweep and the first upward deflection of the NAP trace on an

oscilloscope. NAPs can be recorded from nerve segments resected and placed

on a recording grid, or by using near nerve needle recording and stimulating

electrodes. They are routinely recorded by surface (without skin penetration)

stimulation and recording from humans in clinical nerve conduction laboratories

(“EMG” labs). Cutaneous sensory nerve (e.g., sural nerve at the ankle) NAPs

studied using this method are called “SNAPs” (sensory nerve action potentials).

NAP or SNAP amplitudes are influenced by a number of factors beyond axon

numbers, a topic beyond the scope of this discussion. For example, demyelinat-

ing neuropathies and neuropathies with axonal atrophy are also capable of

influencing both amplitudes and conduction velocities. One critical variable

that the experimentalist (and human EMG lab) can control is the near nerve

temperature during recordings. These should be maintained close to 37�C


(near nerve) in experimental preparations and greater than 32�C (surface skin

temperature) in humans during recordings.

Multifiber techniques are particularly useful in measuring pure motor axon

regeneration. The recordings are made over the endplate of a muscle normally

supplied by the nerve in question. In this technique, the whole nerve trunk

is stimulated with a supramaximal stimulation, but the endplate recording

only captures the potentials arising from activated motor axons. The recorded

potentials are therefore termed CMAPs (compound muscle action potentials) or

“M” waves (Figure 4.3). As with NAPs, there is an excellent correlation between

CMAP amplitude and numbers of viable motor axons that have reconnected to

the endplate zone of the muscle. As above, however, it is not entirely correct to

suggest that only axon numbers influence the CMAP. CMAPs may be dispersed

(spread out) when motor nerves are demyelinated. In rats reared in wire bottom

grid cages (as opposed to plastic sawdust covered cages) significant disortion

of CMAPs is attributed to compression neuropathy of distal hindpaw motor

branches, i.e., “wire cage neuropathy” [804]. The use of these cages should be

discouraged in rodent studies of regeneration.

CMAPs (or NAPs, SNAPs) recorded after stimulation of a nerve trunk distal to

an injury do not immediately disappear despite damage to all axons at the injury

site. Whereas stimulation proximal to the injury site, as expected, fails to recruit

a response, axons distal to a nerve injury remain excitable for several days.

If examined early after injury, these recordings might suggest that the distal

axons are normal. The same recordings carried out 7–10 days later, however, may

fail to evoke any response. Thus, to assess whether distal fibers have been

damaged by a given injury, recordings should therefore be delayed. In a nerve

trunk containing thousands of individual axons, it may be that some axons are

injured but others are not. To then address how extensive the damage is, a CMAP

k n

Figure 4.3 Examples of compound muscle action potentials (CMAPs) recorded

from the interosseous foot muscles of the rat with stimulation of the sciatic nerve

proximally at the sciatic notch (n) or more distally at the knee (k). The smaller

potentials after the main peaks at the end of the tracings are F waves.

(Reproduced with permission from [789].)

67Electrophysiological measures of regeneration

recorded from distal nerve stimulation and measured at the appropriate time

can estimate the proportion of axons damaged in a partial nerve injury. A 50%

reduction in the amplitude of the CMAP from stimulation recorded 7–10 days

after injury suggests that 50% of the motor axons have been irreversibly dam-

aged and rendered inexcitable. Finally, the rate of excitability loss after axon

injury may be delayed in situations where fundamental processes of axonal

degeneration are delayed (e.g., WldS mouse discussed in Chapter 3). All of these

comments hold true whether CMAPs, NAPs, or SNAPs are considered. There are,

however, some special issues concerning CMAPs.

The relationship between the amplitude of a CMAP and the proportion

of injured axons changes with time. If denervation and reinnervation are pro-

longed, collateral sprouting of uninjured distal motor axons occurs and may

appear to repair the CMAP. Residual intact motor axons may sprout to indivi-

dually innervate large numbers of muscle fibers. There may therefore be a rise

in the overall amplitude of the CMAP with little new motor axon regrowth.

With the above caveats, multifiber recordings can provide highly sensitive

serial indices of nerve recovery. These include the time for reappearance of the

first NAP or CMAP after complete injury, the amplitude of the potential at fixed

time points after injury, and the conduction velocity of regenerating axons.

Conduction velocity is reduced in newly regenerating axons because of their

immaturity: their smaller radial caliber, incomplete remyelination, and perhaps

their immature nodal apparatus.

The use of nerve conduction recordings in human disease is not the topic

of this text and will not be covered further here. In rodents, the sciatic nerve is

usually chosen as an accessible, well-characterized site of regeneration studies.

Motor recordings can be made by stimulating the sciatic nerve at the sciatic

notch and popliteal fossa with recordings of CMAPs over the interosseous muscle

endplates of the hindpaw. These are recorded using percutaneous fine platinum

electrodes inserted just beneath the dorsal skin of the paw. Near nerve tempera-

tures are maintained by a subcutaneous subdermal thermistor and overhead

heating lamp. In normal rats and mice, good-quality CMAPs with G1 (active, over

the muscle endplate) and G2 (reference) recording electrodes placed over

mid-dorsal paw (the endplate is approximately halfway along the paw) are

recognized by low stimulation threshold and an initial negative (upward by

North American convention) deflection. An initial downphase instead of upward

deflection indicates poor active recording electrode (G1) placement. After sciatic

nerve injury and before newly growing axons reconnect to the endplate, CMAPs

are unrecordable. Overstimulation at this time, however, can yield artifactual

distant direct muscle stimulation volume potentials. These are recognized by

their higher stimulation threshold and abnormal shape.


Several approaches to estimating numbers of motor units in humans have

been published (motor unit number estimation; MUNE). One approach involves

using an incremental rise in stimulation intensity to identify “all or none” steps

in the recorded CMAP amplitude. Each step represents the addition of a single

motor unit and the amplitude of each step can be averaged. The MUNE is

calculated as the total CMAP divided by the averaged amplitude of a single step

increment [449]. Other approaches involve the use of multiple intramuscular

recording electrodes (macro-EMG) that identify the contribution of individual

motor units to the CMAP [68,97,146].

Needle electromyography (EMG), widely used in human laboratories, provides

a further direct measure of electrophysiological activity in denervated muscle.

It is therefore used to identify regeneration or lack of it. Denervated muscle

fibers develop abnormal spontaneous firing because of fiber depolarization

and redistribution of acetylcholine receptors (AchRs). Calcium accumulates in

sarcoplasmic reticulum luminae. Abnormal spontaneous firing can be easily

recorded by fine caliber needle electrodes during electromyography, a technique

extensively described in other textbooks. Spontaneous potentials are recognized

as “fibrillation potentials” or “positive sharp waves,” two waveform configura-

tions likely representing identical physiological information-spontaneous single

muscle fiber discharges. In electromyographic parlance, fibrillations recorded

with needle electrodes from humans with denervation are sharply contoured

biphasic or triphasic potentials, 1–5ms in duration, 20–200mV in amplitude, and

usually regular (1–30/s) in their firing rate [345] (also see Figure 6.3). Although

their amplitude and firing rates may be quite variable, the density with which

they are encountered by the recording electrode is correlated with the extent

(severity) of denervation. The positive sharp wave, like a fibrillation potential thus

represents a similar single muscle fiber discharge. Positive sharp waves are recog-

nized by an abrupt downward (positive by convention) deflection with a gradual

sloping recovery. Their altered waveform may arise because of the added impact

of the needle electrode on the fiber being recorded from (angle of view, added

injury component) [154,360,498,594]. There is a good correlation between the

density of these potentials and the degree of denervation (complete, degrees of

partial, or none). They disappear early during reinnervation from regeneration.

Like CMAPs, fibrillations and positive sharp waves take several days (depending on

the distance of the injury to the muscle endplate) to appear fully after an injury.

In humans, EMG is also routinely used to assess the number and caliber of

voluntary motor unit potentials that have regrown to muscle or have resisted

injury. As discussed above, collateral sprouting in partial injuries enlarges the size

of these individual motor units. In animal models, recording voluntary activity in

an orderly way is obviously challenging.

69Electrophysiological measures of regeneration

Other electrophysiological techniques are also used to address regeneration.

While more time consuming and requiring greater expertise, rigorous open

approaches to motor unit recording combined with high quality motor unit

force measures provide exquisite analysis of motor function. This approach,

described by the Gordon laboratory [663] accurately identifies numbers of motor

units and their degree of sprouting. Unlike CMAP recordings, they are not

confounded by collateral sprouting or demyelination but instead accurately

address both. The sciatic nerve and tibialis anterior, soleus, medial gastrocne-

mius, or plantaris muscles are exposed and connected to a force transducer

using 2.0 gauge fine-braided silk threads. Twice suprathreshold sciatic nerve

stimulation is applied to generate both the maximal evoked muscle twitch force

and the tetanic force in response to 1, 5, and 21 pulses at 100Hz at a repetition

rate of 0.5–1.0Hz. Single motor unit force is generated by stimulation of ventral

roots L4 or L5 teased into rootlets of 5–10 motor axons and recordings are made

of all or none responses. After sampling approximately 40% of the motor axon

population, the motor unit numbers (numbers of motor axons populating the

whole nerve) are calculated by dividing the whole muscle twitch force by the

mean motor unit twitch force. An excellent correlation exists between motor

unit size enlargement measured this way and the number of histologically

identified collateral sprouts innervating adjacent denervated endplates after

partial denervation.

Some laboratories have published work using F and H wave recordings.

These are long loop potentials generated by stimulating a nerve and awaiting

a potential that travels to either the axon hillock (F wave) or the spinal cord

(H reflex), then back again. They are used to address proximal axon motor and

sensory function, respectively, in neuropathy models. An F wave (Figure 4.3) is a

motor potential, generated by several motor axons in which an applied stimulus,

similar to that used to evoke a CMAP, also travels retrogradely (antidromic

conduction) in the motor system. The retrograde discharge re-excites a pool of

motor axons at the axon hillock, then travels back to the endplate to be recorded

as a late potential. Since F waves represent only a proportion of the excitable

motor axon pool, its usefulness for regeneration work is limited. F waves are

recognized as potentials that are delayed after the CMAP and are smaller in

amplitude.

In many instances, excluding rigorous open electrophysiological approaches,

F waves have been mistaken for H reflexes, a potential generated by stimulating

sensory axons. The H reflex is a long loop monosynaptic reflex akin to the deep

tendon stretch reflex brought out by a neurologist’s reflex hammer. It involves the

recruitment of large myelinated sensory axons. Once activated, these axons synapse

on motor axons in the spinal cord to elicit a response. Specific electrophysiological


characteristics must be present to accurately identify an H reflex: recruitment

by low amplitude long duration currents, its appearance prior to that of the

maximum CMAP, and its decline with higher stimulus intensities.

Evoked potentials from peripheral nerve stimulation and recording over

central structures, like the spine or brain, offer yet another approach toward

addressing reinnervation. The amplitudes of these responses are much more

variable and do not strictly reflect the number of conducting peripheral axons.

This technique is used less frequently.

Target reinnervation

Regeneration of peripheral nerves can be assessed by measuring the

onset and extent of target tissue reinnervation. Examples include reinnervation

of the epidermis by unmyelinated sensory axons and neuromuscular junction

reinnervation.

Consensus guidelines have been developed for processing and interpreting

human biopsy samples of the epidermis [377]. Most laboratories have used the

axon marker PGP 9.5 in thick (30–50 micron thickness) sections and calculate the

number of axons per epidermal length. Density measurements, the number

of axons per unit area of the epidermis, can also be calculated. Epidermal axons

can be identified by fine axon processes, labeled by PGP 9.5, extending upward

90 degrees to the axis of the skin. The classification of a “single” axon may require

that it crosses the dermal epidermal junction. Epidermal axons thus originate

from longitudinal axons coursing at, or just beneath, the dermal-epidermal

junction. Other measures applied to epidermal axons include: fiber length,

numbers of degenerating profiles (“bulbar” degeneration with profiles resembling

degenerating axon endbulbs), and dermal measures of sweat gland innervation.

Functional reinnervation of sweat glands can be measured by the subcutaneous

injection of pilcarpine (5mg/kg) and measuring sweat droplet impressions

made in a silicone mold of the hindpaw [499]. One sweat droplet identifies one

sweat gland.

For motor axon reinnervation of neuromuscular junctions, immunohisto-

chemistry can be used. For example, it is possible to double label axons (using

neurofilament or ChAT) with a bunagarotoxin (or another label) used to label

acetylcholine receptors at the endplate (see Figure 2.16). The number of total,

innervated, or denervated endplates (those with or without an axon, respectively)

can be counted. Sprouting of intact axons to reinnervate adjacent denervated

endplates can also be classified, i.e., (i) pre-terminal in which a sprout arises prior

to the normal neuromuscular junction of the axon; (ii) ultra-terminal where a

sprout arises at the terminal; and (iii) nodal sprouting in which the sprout arises

71Target reinnervation

from the first node of Ranvier proximally in the parent axon. These measures

help to characterize the type and extent of the regenerative sprouting [664].

Retrograde, anterograde labeling

Regrograde labeling involves the application of an indicator substance,

usually a fluorochrome, to the zone of regenerating axons. The indicator is taken

up by axons at the injection site and then is retrogradely transported to the

parent cell bodies. The numbers of cell bodies, in turn, reflect the number of

neurons that project axons to the injection site. The injection site may be in the

target tissue or in the nerve itself at a fixed distance from the injury. The most

common indicators are fluorogold (FG), fast blue, mini-ruby, fluororuby (dextran

tetramethylrhodamine), fluoro-emerald, diamidino yellow, HRP (horseradish per-

oxidase), lectins (wheatgerm-agglutinin, concanavalin-A, and others), biocytin/

neurobiotin, latex beads, rhodamine-isothiocyanate, carbocyanines such as

DiI (1,10-dioctadecyl-3,3,30,30-tetramethylindocarboncyanine perchlorate), and others

[351]. Since the concentrations and approaches vary with each indicator, it is

important to refer to published material on their use. FG requires several days

(3–4) to be retrogradely transported and label cell bodies. Fast blue (2%), FG (2%),

mini ruby (10% solution of dextran, tetramethylrhodamine, and biotin), fluor-

oruby (10% solution of dextran, tetramethylrhodamine) all labeled a full and

comparable complement of motor neurons by 1 week when applied to freshly

cut adult rat medial gastrocnemius nerve. The labels were applied for 2h in an

otherwise sealed (except to the proximal nerve stump) polyethylene tube.

Fluoro-emerald (10% solution of dextran, fluorescein) labeled about 10% fewer

neurons. By 4 weeks, fast blue and FG labeling persisted, whereas other labels

had declines in their expression. Only fast blue stained a near complete majority

of motor neurons out to 12 weeks and 24 weeks [511]. For comprehensive reviews

of tracing methods see Kobbert et al. [351] and Tonge et al. [684] (Figure 4.4).

There are additional pitfalls using retrograde labeling. It is important to ensure

that the indicator does not spread significantly beyond the injection site where it

might label other axons. In some conditions, axons may not take up the label (e.g.,

FG labeling can be optimized if the nerve trunk is cut and dipped into the source

of the indicator). On occasion, uptake might be inexplicably poor and not visu-

alized. Sometimes perineuronal satellite cells acquire the indicator and their

labeling might obscure accurate counts of labeled neurons. There are also caveats

to performing unbiased three-dimensional neuron counts, not discussed here.

Physical or optical dissector counting methods are the techniques of choice [117].

The advantage of retrograde labeling is that it strictly identifies the number of

parent neurons that contribute to new growth. This is important because it is


Figure 4.4 Images of DRGs retrogradely labeled with fluorogold (FG, blue) after a

sural nerve injury. The sections are colabelled with an antibody against neurofilament

(red). (Image taken by X.-Q. Li, Zochodne laboratory.) See color plate section.

73Retrograde, anterograde labeling

possible that a single neuron and parent axon might generate several regrowing

axons distal to an injury site. If only sections of nerve distal to the injury site are

examined, it might be incorrect to surmise that all the axons observed are from

independent parents. Since many injuries are also associated with retrograde

loss of parent neurons, tracing approaches can help to identify this kind of loss.

Thus, a pool of parent neurons that decreases in number after injury indicates

specific retrograde loss of neurons supplying the nerve. It is also possible to

examine abnormal innervation patterns by using two retrograde labels that

fluoresce at differing wavelengths. For example, motor neurons that inappropri-

ately send axons down incorrect cutaneous branches during reinnervation can

be identified by applying a tracer to the incorrect cutaneous nerve and another

to the correct motor branch [7]. Incorrectly projecting neurons may be distin-

guished from those normally reinnervating their appropriate nerve.

Anterograde labeling is also available to examine the distance of nerve

regrowth. In one approach, neurobiotin tracer was injected into freshly cut nerve

approximately 3mm proximal to a previous zone of crush carried out 3 days

earlier. The samples were harvested for labeling of the distal extent of axon

growth 5 hours later (streptavidin cyanine 3 fluorescent marker) [716]. It is also

possible to inject [35S] methionine into the area of the cell body (e.g., anterior

horn) and to measure radioactivity at specific distances from the injury site [73].

Pinch test

The pinch test is an older and less frequently used technique that

examines the distance traveled by outgrowing nerves after a time frame is allotted

for regeneration to begin [320,418]. Themethod involves exposing an injured nerve

trunk, for example, in a rodent that is lightly anesthetized. Starting distally, a pair

of fine tipped forceps is used to gently squeeze the nerve until it evokes either

a twitch from a muscle remote from the injury site (e.g., abdominal muscle) or a

vocalization. The distance between this site and the injury site (previously marked

with a suture or ink) is measured with sensitive calipers. The resulting measure-

ment provides the distance traveled by the fastest regenerating sensory axons.

While widely used in the past, the pinch test does have significant disadvan-

tages. The site eliciting twitch or vocalization cannot be strictly ascribed as

the location of the fastest growing axons. By pinching a nerve, even gently, the

distortion of the nerve trunk may be capable of generating abnormal retrograde

discharges (basis of the muscle twitch or vocalization response) of axons in

a variable zone both proximal and distal to the regenerating axon tips. Abnormal

architecture of the nerve after injury might complicate identifying which exact

axons are sensitive to the forceps. Theoretically, the pinch test only measures


regrowth of sensory axons. It may also be the case that early regenerating sprouts

are less electrically excitable and that only more proximal portions of new axons

are capable of generating discharges. Accurate distance measures, particularly

after short regenerating intervals, may be difficult to ascertain. Finally, the test

deliberately distorts nerve trunks that require gentle handling for high quality

histological studies. Pinch testing may thereby ruin histological studies.

Functional sensory recovery

Behavioral tests of sensory recovery provide an important index of

peripheral nerve regeneration. All modalities of sensation can be tested in

humans by standard neurological examination: sensitivity to light touch, pin-

prick, cold, vibration, and position. Quantitative approaches (known as QST for

quantitative sensory testing) are also available. The most rigorous QST aloga-

rithms, as developed by Dyck and colleagues, involve a blinded (subject) forced

choice test approach. Details are reviewed elsewhere [158,161,623].

It is more challenging to assess all functional forms of sensory recovery

in rodent models. There are protocols from the experimental pain literature

describing testing for thermal sensation (thermal hyperalgesia) and mechanical

sensitivity (allodynia or inappropriate pain behavior to a nonpainful stimulus).

The recovery of sensation to a thermal stimulus after a peripheral nerve injury is

measured by the latency to paw withdrawal from heat. A testing apparatus,

designed by Hargreaves, provides a non-noxious heat stimulus and measures

the time it takes for the animal to withdraw its paw [253]. Recovery of thermal

sensation can be complex after nerve injuries. Early after injury, the withdrawal

latency is prolonged or absent indicating sensory loss. If there is no response, the

examiner should discontinue the heat stimulus in order to prevent injury. A heat

stimulus, however, may not be confined to the nerve of interest. Since a signifi-

cant volume of the hindpaw is heated, adjacent intact territories may also be

stimulated, leading to withdrawal despite ongoing sensory loss in the index

nerve. Collateral sprouting can extend the territory supplied by adjacent nerves,

into the zone previously innervated by the injured nerve. Thus an apparently

recovered thermal withdrawal may arise from collateral branches of intact

nerves, not from recovery of the original lesion. Similarly, partial injuries

can be associated with hyperalgesia (increased sensitivity) by intact adjacent

nerves. In a given injury, therefore, understanding the contributions of all of

these confounding factors may be difficult. Serial measurements can help to

resolve some of this difficulty. These can identify, for example, an initial period

of sensory loss followed by hyperalgesia, then finally a return to normal sensa-

tion. Finally, two technical considerations are important. First, animals should

75Functional sensory recovery

undergo several sessions of testing to accustomize them to the experiment.

Second, when testing unilateral injuries, the injured side should ideally be

compared to the contralateral intact side and to the pre-injured state.

Recovery of mechanical sensation may be studied by the response of the rat or

mouse hindpaw to monofilaments. Known as von Frey hairs, these can be lightly

applied to the plantar surface of the hindpaw of a rodent while it is standing

over meshed flooring. The animals do not require handling, but should be

acclimatized to their housing and unique flooring. Other stress factors should

be eliminated and training should be carried out before the experiment. The

filament is inserted though the mesh and is pressed lightly against the paw until

it begins to bend. Bending forces are calibrated for a range of filaments.

For example, in humans a 10 gram monofilament (Semmes–Weinstein monofila-

ment) applied to the surface of the toe is used to quantitate the severity of

sensory loss from a neuropathy. In rodents, the withdrawal of the hindpaw to

each monofilament can be graded by the examiner. For example, each hindlimb

can be tested six times, and the number of times there is foot withdrawal for

each stimulus is recorded. Allodynia is an exaggerated response to an innocuous

stimulus such as a monofilament with a smaller bending force that does not

usually provoke withdrawal. After peripheral nerve injury, allodynia, like thermal

hyperalgesia, can develop because of collateral sprouting. Overall, however, mono-

filaments test a much more discrete localized zone of sensation than heat stimuli.

As reinnervation proceeds and sensation returns, withdrawal occurs first to coarse

monofilaments then to finer monofilaments.

Galtrey and Fawcett [208] studied median and ulnar nerve injuries (crush or

transection) in the forelimb of the rat. Results of sensory testing in normal

control limbs and in limbs 15 weeks after nerve injuries were provided. For the

forelimb, the authors analyzed electronic von Frey tactile stimuli. Cold sensory

testing was assessed with an ice probe made from a 1.5ml microcentrifuge tube

placed in freezing water. The mechanical stimuli provided graded force stimuli,

whereas the latency to withdrawal from cold was timed.

An alternative test of sensory reinnervation involves antidromic signaling.

“Antidromic” refers to the propagation of an impulse along a sensory nerve from

proximal to distal, the opposite of normal sensory transmission. The property

underlies what is known as “neurogenic inflammation” in which peptides

released from sensory axons are released in the periphery and cause both vaso-

dilation and plasma extravasation (plasma leakage from capillaries). This feature

can then be exploited to determine whether sensory axons have reinnervated

the blood vessels and associated mast cells of the skin. Low frequency stimula-

tion of the proximal sensory nerve thus activates distal sensory axons and causes

them to release the neuropeptides (SP and CGRP among others) that mediate


antidromic plasma extravasation. If the animal is concurrently perfused with

Evan’s blue, blue discoloration will develop at the sites of plasma extravasation

in the innervated skin. The technique thereby delineates innervated from dener-

vated skin [45,305,354]. A related approach measures rises in skin blood flow in

response to antidromic stimulation to detect reinnervation.

“Metabosensitive” afferent responses are recordings from sensory axons within

muscles in response to fatiguing contractions, local infusions of potassium chlor-

ide, or infusions of lactic acid. When muscles are reinnervated, recovery of these

afferent responses can be recorded in filament bundles of the proximal nerve

trunk [9].

Functional motor recovery

There are a number of techniques available to test functional motor

recovery in rodent models or humans. In rats and mice, quantitative measures

of hindpaw function are widely used. Walking track analysis was described by

deMedinaceli and colleagues [131]. The technique involves placement of the rodent

hindpaw in photographic developer solution. The gait of the rat is then tested by

having them walk on unexposed X-ray films where images of their hindprints

will be made. Hindprints can also be made by placing the paw in other indicators

such as India ink. Measurements aremade from the hindprints to calculate a sciatic

functional index (SFI) given as a percentage. The formula for calculating the SFI is:

SFI ¼ ½ðETOF�NTOFÞ=NTOFþðNPL� EPLÞ=EPLþ ðETS�NTSÞ=NTSþðEIT�NITÞ=NIT��220=4 (all in mm) where the following are measured from the hindpaw imprint:

ETOF ¼ measure from the tip of the experimental foot to the tip of the

following contralateral toes

NTOF ¼ distance from the tip of the normal contralateral foot to the tip of the

following experimental foot

EPL ¼ experimental (E) paw footprint length

NPL ¼ normal (N) paw footprint length

EIT ¼ experimental paw distance between the center of the second and

fouth toes

NIT ¼ as above for the normal paw

ETS ¼ maximum experimental paw toe spread between first and fifth toes

NTS ¼ as above in normal paw

Zero percent (�11%) represents normal function and �100% represents complete

nerve interruption. An example of footprint analysis is given in Figure 4.5.

Further reports have described modifications and variations of the SFI,

not reviewed here. Their advantages over the original scale may be marginal.

For median and ulnar lesions of the forelimb in the rat, Galtrey and Fawcett [208]

77Functional motor recovery

used a modification of SFI described by Ozmen et al. [540] termed pawprint

analysis. Toe spread (intermediate ITS and widest TS) and print length are the

critical measurements elicited. Overall, walking track analysis and pawprint

analysis both provide crucial indices of recovery from nerve injury. It is import-

ant to recognize that it does not simply reflect motor recovery. Gait is a complex

behavior that involves coordinated motor output linked to ongoing sensory

afferent feedback. Sensory loss has a significant impact on walking. The most

suitable tests for evaluating functional recovery after a peripheral nerve injury

depend on the site and type of lesion. For example, Eberhardt and colleagues

[162] used beam walking trials to test for recovery from femoral nerve injuries

that denervate the more proximal quadriceps muscle. Accurate beam walking

requires both intact motor and intact sensory function.

40 mm = ET OF

35 mm

30 mm

40 mm

NTS

NPL

Experimental

50 mm

75 mm = NT OF

ETS

EIT

EPL

NIT

Normal

Figure 4.5 Examples of walking prints for analysis from rats with an injured sciatic

nerve (left) compared to the contralateral intact nerve (right). (Reproduced with

permission from [131].)


Sensitive strain gauges are used to evaluate the resistance of flexed front or

hindlimb digits to pulling. The strength of digit flexion is measured as the

tension at which the animal is able to hold on to a wire grid before releasing.

A caveat is that longstanding denervation of the foot may generate flexion

contractures of the paw. This fixed resistance must be controlled for in the

measurements. Data for forelimb grip strength in rats are available [208], and

the technique can also be applied to mice.

Specific forelimb tests of motor function have also been described to test

recovery from median and ulnar injuries in rats [208]. The staircase test is a task

that requires climbing and reaching for food pellets. The number of stairs

climbed and the amount of pellet retrieval can be quantitated. The horizontal

ladder test for rats involves walking along horizontal rungs to measure misses,

slips, corrections, and partial placement of the forepaws on the rungs [208,463].

Most of the behavioral tests described rely on a training, acclimitization period

before lesions can be studied. Like the SFI, forelimb motor behavior tests also

depend on intact sensory feedback.

In vitro measures of nerve regeneration

An analysis of the outgrowth from harvested and cultured neurons

can provide critical insights into regeneration. The high resolution provided by

in vitro studies permits assessment of axon behavior, especially over short periods

of time. Growth cone morphology, direction, and speed can be visualized and

tracked. Similarly, retraction or withdrawal can be rapidly identified using video

recording. Finally, in vitro approaches allow new molecular pathways to be

rigorously explored before in vivo experiments. The disadvantages are obvious.

In vitro studies do not reflect actual axon growth and navigation in complex

multicellular tissues. Removal from a live animal pre-injures neurons and they

are subsequently required to grow in artificial conditions. Since embryonic

neurons grow at rates much faster than those of adult neurons, interpretation

of their behavior should be guarded. Similarly, “neuron-like” cells such as PC12

pheochromocytoma cells that are easily grown allow exploration of early growth

cone behavior. Findings in these systems, however, require confirmation in adult

neurons prior to extending the approach in vivo.

By convention, outgrowing processes from neurons in vitro are labeled

“neurites” because of the difficulty classifying them as axons or dendrites.

Neurons plated at low density allow individual cells and growth cones to be

observed. Different types of neurite outgrowth are important to examine. The

percentage of neurons with neurites extending from them (process bearing

neurons) reflect neurite initiation. Neurite outgrowth may be evaluated by

79In vitro measures of nerve regeneration

measuring “total” neurite length, the sum of the length of all neurites. Other

measures include the mean neurite length and length of the longest neurite.

Gavazzi et al. [212] noted that neurite initiation and the total neurite length

were trophic factor dependent but that the length of the longest neurite was not.

Facilitated neurite outgrowth can be observed when specific growth factors are

combined with basement membrane substrates (e.g., laminin and GDNF for IB-4

neurons [698]). These findings reinforce the caveat that measurements of neurite

behavior depend heavily on specific culture conditions, including the substrate

and whether specific growth factors have been added to the media.

A number of specific neurite counting approaches have been described.

Mearow and colleagues [698] assessed neurite outgrowth at 24h after plating

and individual tracings of neurons labeled with bIII tubulin were made using a

camera lucida technique. Neurite initiation was defined as the percentage of

total neurons having neurites twice the length of their cell body diameter. Only

unambiguous neurons were chosen. A technique known as Sholl analysis was

used to measure the number of intersection points of outgrowing neurons across

20 micron concentric circles radiating from the cell body. At least 50 neurons

were traced for each condition studied. Most investigators also recommend that

data sets include cells from several animals and that assays are done on different

harvested plates and days.

Compartmented or “Campenot” sympathetic neuron culture dishes were

described by Robert Campenot at the University of Alberta [89] (see review

[705]). Neurons are harvested and placed in the center compartment of a culture

dish. The dish is divided into compartments by a Teflon divider that adheres to

its base with silicone grease. Cytosine arabinoside is added to kill nonneuronal

cells and NGF is added to the neurons. Neurites grow out from the perikarya

in the central compartment, under the silicone grease barriers and enter the

side compartments. As a result, the center compartment contains the perikarya

and the side chambers contain axons so that the compartments isolate axons

from the perikarya. The technique has been a powerful tool to understand the

behavior of axons independent of their cell body. Differences in signaling,

protein synthesis, or lipid synthesis have been examined (see [344]).

Explants

Explants are resected whole DRG or sympathetic ganglia placed in

a media. From these cultured ganglia axons emerge and can be examined.

Explants therefore retain the multicellular milieu of a whole ganglia and are

easier to examine than dissociated single neurons. The original investigations of

Levi-Montalcini and Hamburger leading to the discovery of NGF used sympathetic


ganglia explants to evaluate axon outgrowth. More recently, several innovative

explant approaches have been used to study regeneration. Torigoe et al. [687]

placed transected nerve stumps connected to their ganglia between two sheets

of plastic film and examined outgrowth of axons and SCs for up to 4 days later.

Axons were identified by silver nitrate impregnation and SCs by S-100 immuno-

histochemistry. “Reactive” SCs from a previously injured contralateral nerve

could be added to enhance outgrowth. In a related approach proximal stumps

and DRGs of transected intercostal nerves were placed into shallow gels of

Type I collagen [381,685]. Axon outgrowth from the cut end was identified using

PGP 9.5 labeling and dark-ground illumination [381].

A frog preparation used resected sciatic nerves and their DRGs placed into a

culture system resembling that described by Campenot [685]. The connected

ganglionic and axonal components were isolated in different chambers and

the distal axon was crushed. To determine later axon outgrowth distance,

L-[4,5-3H]leucine was added into the ganglion compartment and axonal transport

was allowed to proceed for 24h. The labeled regenerative front was identified

by autoradiography. Advantages were that the preparation could be carried out

at room temperature and survived without serum or growth factors for at least

12 days. A similar preparation was described using mice with intercostal nerves

or sciatic nerves and their DRGs attached (L4,5 for the sciatic preparation).

The mammalian explants, however, were not viable for as long as those of frogs.

Collateral sprouting

Collateral sprouting differs from regenerative sprouting because it

requires intact axons to sprout branches into denervated sites. Thus, no injury

response of the parent axon is involved. Since many types of nerve damage

involve irretrievable loss of the entire neuron tree, collateral sprouting from

intact residual neurons can be essential to restore aspects of function. Diamond

and colleagues [139,141,149,283,301,508,677] have described an elegant approach

toward assessing collateral sprouting in denervated skin. The technique involved

preserving a sensory nerve adjacent to denervated skin. Thus, the skin of one side

of the dorsum of a rat is denervated by sectioning four dorsal cutaneous nerves

rostral and caudal to T13, the lateral branch of T13 and four lateral cutaneous

nerves supplying adjacent flank skin. A single nerve territory T13, adjacent

to these resections is preserved and mapping of the skin area retaining sensation

in its innervation territory can be carried out. This map includes sensation of

the skin to pinch (mechanosensitive Ad fibers), and to a 60�C heat probe (heat

nociceptive C fibers). In lightly anesthetized rats, the pinch and heat sensations

can be categorized as present or absent depending on a skin twitch of the

81Collateral sprouting

underlying cutaneous trunci muscle. The map may identify hyperalgesia in some

experiments. Sensation to light touch is identified by recording impulses from

large myelinated Ab axons of the dorsal cutaneous nerve in response to brushing

the skin with a fine bristle. Collateral reinnervation involves expansion of

sensitivity over time from T13 into surrounding areas that have been denervated.

A related approach was used to document the area of sensory loss after sural

nerve biopsies in patients. Using light touch and pinprick, the areas of sural nerve

sensory loss or sensory attenuation were mapped on the side of the foot and ankle.

Serial measurements were used to determine how recovery occurred. Over time, a

symmetrical contraction, from the edges in, of the denervated zone was identified

indicating collateral reinnervation from adjacent intact nerves [678] (Figure 4.6).

6 months post-biopsy

18 months post-biops y

Figure 4.6 Maps of the areas of sensory loss from the side of the foot in a patient

that had undergone a diagnostic nerve biopsy with complete resection of the sural nerve

at the side of the lower leg. The black areas represent complete loss of sensation to

pinprick and the hatched areas outside of this represent partial loss of pinprick sensation.

Over time (bottom panel) the area of sensory loss has gradually retracted, accounted

for by collateral reinnervation from intact territories adjacent to the denervated skin.

(Illustration by M. Theriault and reproduced with permission from [678].)


Summary

Rigorous assessment of peripheral nerve regeneration involves unique

techniques and interpretation. In vivo, morphological assessment of axons at fixed

distances beyond an injury zone using epon- or plastic-embedded semithin and

thin sections prepared in iso-osmolar fixative is a gold standard approach.

It provides information about the number, density, and maturity of outgrowing

myelinated axons under lightmicroscopy and unmyelinated axons using electron

microscopy. Careful immunohistochemical approaches that analyze individual

axons at early time points during regeneration can address the amount and

distance of early regrowth. Retrograde techniques are similarly very important

in addressing whether retrograde loss of parent neurons or multiple daughter

axons have emerged. Electrophysiology allows serial assessment of regeneration

of motor or sensory axons. In vitro approaches including dissociated neuron

cultures or explants allow testing of early molecular determinants of growth

cone behavior. Finally, behavioral assays of motor and sensory function are

essential in addressing the success of functional recovery. New high resolution

imaging techniques capable of identifying the extent of axon outgrowth in vivo

offer exciting future prospects.

Suggested reading

de Medinaceli, L., Freed, W. J., & Wyatt, R. J. (1982). An index of the functional

condition of rat sciatic nerve based on measurements made from walking tracks.

Experimental Neurology, 77, 634–643 [131].

Diamond, J., Gloster, A., & Kitchener, P. (1992). Regulation of the sensory innervation

of skin: trophic control of collateral sprouting. In Sensory Neurons: Diversity,

Development and Plasticity, ed. S. A. Scott. Oxford, UK: Oxford University Press,

pp. 309–332 [140].

Tonge, D., Edstrom, A., & Ekstrom, P. (1998). Use of explant cultures of peripheral

nerves of adult vertebrates to study axonal regeneration in vitro. Program

Neurobiology, 54 (4), 459–480 [685].

Appendix 4a

Protocol for fixation and processing of nerves for semithin and thin sections

(adapted from Dyck Laboratory)

Fixation: Place harvested tissue on a stick (e.g., tied at one end gently to a toothpick)

and store in small vial (i.e., 1–3ml) of 2.5% glutaraldehyde buffered by 0.025M

cacodylate (pH 7.4) solution. One end of the stick may be tied with a suture

for orientation. Fix for 6 hours to overnight. Wash the nerve for 30 minutes in

6 changes (5 minutes each) of fresh solutions of 0.15M cacodylate buffer (pH 7.4).

Tissues may be cut afterward in smaller pieces prior to osmification (cut distal

83Summary

angled and proximal straight with a fresh, sharp scapel blade). Tissues in 0.15M

cacodylate buffer can be stored for long periods.

Osmification: Prepare osmium by breaking 1 vial of osmium tetroxide into 25ml

of ddH2O. Let osmium dissolve for 3–5 days. Keep cool. Do not allow solution to

turn black (should be yellow to light brown). Dissolved osmium may be stored in

the refrigerator for approximately 2 weeks or until it darkens (dark brown/black).

Osmify samples after the rinse step above. Mix 2% osmium tetroxide in

0.12M cacodylate for 2.5 hours, then wash for 30 minutes (6�5 minutes) in

0.15M cacodylate.

Dehydration: Pipette off storage buffer and rinse with 70% ETOH. Maintain in this

for 8 minutes. Pipette off 70% ETOH and rinse with 80% ETOH. Maintain this for

8 minutes. Repeat rinses, each for 8 minutes with 95%, 100% (absolute ethanol),

repeat 100%, propylene oxide, and repeat propylene oxide.

Infiltration: Mix epon/propylene oxide solution (50% epon/50% propylene oxide;

see below). Add �1ml of 50/50 mix to each sample and allow to infiltrate over

night in a fume hood (�12 hours) and to allow the propylene oxide to evaporate.

Embedding: Two types of epon mixture are used. Mixture A is Epon 812 (Jembed,

Canemco, St. Laurent Quebec) 31ml with DDSA (Dodenyl succinic anhydride;

Canemco) 50ml. The second, Mixture B, is Epon 812 100ml and NMA (Nadic

methyl anhydride; Canemco) 89ml. The solutions are kept for one hour at

room temperature for use. All glassware should be dried thoroughly. In a 50ml

graduated cylinder 15ml of Mixture A and 35ml of Mixture B is then placed in a

beaker. Next 1ml of DMP-30 (Tri-[dimethylaminomethyl] phenol; Canemco)

is added, the mixture covered with foil and mixed for 10 minutes with an egg

beater. Let epon sit until all bubbles settle out. Note: It is only worth preparing

mixes A and B in advance when doing large numbers of samples over the space

of a week. Otherwise it is simplest to prepare mixes the day of use. The tissue

is dipped through two changes of epon and then embedded in a block mold.

Epon should be used within 4 hours of mixing to ensure freshness. Fill each mold

(e.g., 6 sample mold) with 1mm layer of epon, then place the sample in the

narrow end of the mold. The specimen should be oriented in the mold so that the

end of interest faces the narrow face where sectioning will occur. The mold

should be filled generously.

Baking/Curing: Put filled molds in an oven preheated to 45�C. Check the samples after

12 hours for alignment, then bake for approximately 24 hours. Next increase

temperature to 65�C on second day and allow another 24 hours or until blocks are

sufficiently hardened. Allow the blocks to cool, then remove from molds. Do not

overcook the blocks. Appropriate hardening lock is hard enough when you can

lightly mark it with a finger nail by pressing firmly. (For all procedures, technicians

should be gloved with appropriate eyeware, clothing protection during

preparations and procedures should be carried out in a functioning fumehood.)

Sectioning: Trimmed narrow faces of epon blocks can be sectioned, using an

ultramicrotome, at 0.5–1.0 microns in thickness. Sections are placed on glass

slides, stained with toluidine blue, and coverslipped.


5

Early regenerative events

In this chapter I discuss the earliest events of peripheral nerve

regeneration. What will emerge from this information is a remarkable panoply

of cells and molecules that come together for regrowth to occur. Considered

separately, the anatomical events, the signaling interactions, and how they

influence each other speak to a coordinated beauty. While the anatomical, or

cellular features of early nerve regrowth have been described for many years,

the molecular cascades involved are only now being considered. Some of this

information, such as growth cone signaling and guidance, has not been exam-

ined in the specific situation of peripheral nerve regrowth. Nevertheless,

there is strong evidence that it eventually will be identified as important

in peripheral neurons as in the experimental and central systems studied to

date. Thus, some of the material presented relates to the general neurobiology

of the growth cone and axon growth, only summarized here. Seen as the

forest, rather than the trees, however, it is a critical part of the peripheral nerve

regeneration story.

Early axonal events

Immediate responses to injury

When an axon is physically disrupted, a rapid series of molecular events

ensues over the next seconds, minutes, and hours. Injured axons undergo rapid

depolarization, generating immediate retrogradely propagated “injury poten-

tials.” These critical signals, perhaps associated with a calcium wave, from distal

axon to proximal portions of the neuron, may be the first to initiate a shift

in its properties. That is, the neuron begins to relinquish its role as a stable

transmitter to one primed for regeneration.

85

Transected axons quickly (within hours) seal their ends after transection to

prevent further egress of axoplasm. Rapid axon repair requires new membrane

synthesis either by local means or from transport along the length of the axon.

Not all injuries involve disruption or transection of the axon membrane. Crush

or stretch inflict similar types of damage to individual axons without actual

physical separation of the axon membrane. Clearly, more than resealing is

required to reverse the immediate consequences of injury. This may involve

the supporting lattice of the axon, for example, the neurofilament and micro-

tubule networks. Shattered cytoskeletal proteins may disrupt axoplasmic trans-

port and membrane activity that influence recovery. Along these lines, axons

from transgenic animals lacking neurofilaments may actually be harder to

crush [176]. Whether physical transection or damage from crush or stretch

occurs, there is irreversible and irreparable disruption. Downstream axonal

Wallerian-like degeneration ensues even if the proximal and distal stumps of

the individual axons remain connected. The trigger for axon demise may be

transient, such as the calcium wave described, but may be sufficient to activate

the machinery for active Wallerian degeneration (see Chapter 3). When periph-

eral nerve trunks are transected, retrograde degeneration to the first node of

Ranvier in the proximal axon also occurs. The retrograde reaction is associated

with all of the changes described in distal Wallerian-like degeneration

including SC activation and macrophage invasion. Consequently, new axons that

grow from proximal stumps are required to make up some lost time to arrive

at the injury site!

The distal ends of transected axons undergo ultrastructural reorganization

early after injury. In the first few hours after transection, microtubules are

first disassembled, then reorganized to form “traps” where vesicles accumulate

[171]. As will be discussed below, microtubules are polar longitudinal struc-

tures with a þ and – end. In intact axons, the microtubules point in one

direction with the þ end directed distally. Following injury, however, their

orientation changes. At 50–150mm proximal from the resealed axon tip at the

injury site, þ ends of microtubules reorientate and point to one another

instead of pointing in one direction. This is called a positive “trap” and it

is associated with collections of arrested anterogradely transported vesicles.

These vesicles help to form the growth cone organizing complex (GCOC).

Subsequent growth cone formation thus requires that the GCOC fuses antero-

gradely transported vesicles to provide membrane material. Golgi-derived ves-

icles are also associated with the GCOC. A separate minus “trap” found more

distally in the axon is associated with – ends of misoriented microtubules.

Minus “traps” help to segregate “old” membrane material away from the

positive “trap.”

86 Early regenerative events

Table 5.1 Some regeneration related proteins

Protein Alternative name Role

AIP Actin interacting protein-1 Helps cofilin to depolymerize F-actin

Akt Protein kinase B (PKB) Signaling and outgrowth signal from PI3K

APC Adenomatous polyposis coli Tumor suppressor protein, interacts with

microtubules

Arp 2/3 Actin-related protein Promotes actin formation

Cofilin Actin depolymerizing factor (ADF) Involved in both polymerization and

depolymerization of actin

CDC42 Promotes growth cone extension

CRMPs Collapsin-response mediator

proteins

CRMPs 2,4 are known to inhibit outgrowth

Dhh Desert hedgehog protein Involved in nerve fasciculation

GAP43/B50 Growth-associated protein 43/B50 Actin interacting protein

GSK3b Glycogen synthase kinase beta Suppresses axon formation

Katinin Microtubule severing

IL-6 Interleukin-6 Promotes regeneration overmyelin inhibitors

LIM kinase Inactivates cofilin and blocks growth cone

advance

mTOR Mammalian target of rapamycin ?branching (promotes dendrite arborization

through PI3K)

NRG Neuregulin Signal from axon to SC

NF-kB Nuclear transcription factor

kappa B

Survival transcription factor that is

upregulated and translocated to nucleus

after axotomy

PAK P21-activated kinase Effector of Rac to facilitate growth cone

extension

PI3K Phosphatidylinositol-3 kinase Survival and outgrowth pathway

PKC Protein kinase C Intracellular signaling molecule (multiple

roles, subtypes)

Profilin Influences actin polymerization; complex

actions

PIP2 Inhibits cofilin actin binding

PTEN Phosphatase and tensin

homolog deleted on

chromosome ten

Suppresses PI3K signaling and increases

unphosphorylated GSK3b mediated

growth cone collapse

RAC1 Promotes growth cone extension

ROK RHO kinase Promotes growth cone collapse

RHOA Promotes growth cone collapse

Sema3A Semaphorin 3A Promotes growth cone collapse

Slingshot Activates (dephosphorylates) cofilin and

promotes outgrowth

ZAS3 Zinc finger transcription factor that

competes with NF-kB

87Early axonal events

Axonal endbulbs and their constituents

Later following axon transection more dramatic changes in the structure

of proximal axons develop (Figure 5.1). Within 24–48 hours after injury the

proximal ends of damaged axons swell, creating structures known as axon end-

bulbs or boutons. While it is possible that endbulbs develop from plus “traps” they

develop later than the microtubular changes described above. First described by

Cajal [572] as “sterile clubs” or necrotic profiles, endbulbs are larger structures.

They are also to be distinguished from “clubs of growth or buds” that represent

new growth cones. Unlike a normal axon terminal, endbulbs exhibit defective

“turnaround” of material from antegrade to retrograde transport. Ongoing

anterograde axoplasmic transport continues unabated and endbulbs form in

swollen axons that contain axoplasmic material, depolymerized neurofilaments,

but also apparently functional molecules [153,192,193,393,572,788,800]. Some

endbulbs have complex arborizations (helicoids or ring like) of neurofilaments

within them that Cajal thought were precursors of outgrowth, and called “struc-

tures of Perroncito” named after a nineteenth-century anatomist. Axonal end-

bulbs also persist, especially if regeneration is frustrated, and they can be

identified indefinitely after some injuries. For example, in the CNS, they have

been identified in brain white matter for years after axonal shear injuries

from trauma.

Endbulbs are more than simple pathological curiosities. Besides structural

proteins, they also house functional peptides and enzymes. Cajal suggested that

endbulbs undergo slow dissolution and that intact viable axons of the proximal

stump withdraw contact from them. Once no longer confined to an axon inter-

ior, molecules released from the endbulbs are free to influence other cells in the

vicinity of the damaged nerve trunk. Thus, the neuropeptide CGRP (calcitonin

gene-related peptide) is probably slowly released from endbulbs, and when in

the extracellular space behaves as a potent vasodilator of local microvessels

[784,788]; it can serve as a mitogen for Schwann cells [111] and it may act

synergistically with SP (Substance P) to facilitate local ectopic discharges that

initiate neuropathic pain (Figure 5.2). Other peptides have been identified in

axonal endbulbs including SP, NPY, and galanin [788]. Furthermore, endbulbs are

sites of aberrant sodium channel accumulation, which are also linked to the

generation of ectopic axon impulses and pain [135]. Endbulbs, however, also

express m opioid receptors, normally associated with the dorsal horn of the spinal

cord. As in the spinal cord, peripheral m opioid receptors in endbulbs are capable

of dampening pain signaling [694] (Figure 5.2).

Axonal endbulbs also accumulate isoforms of nitric oxide synthase (NOS), both

eNOS (endothelial NOS) and nNOS (neuronal NOS) [393,394,800]. While nNOS

has long been described in neurons and is upregulated after injury, eNOS was


once thought to be exclusive to endothelial cells. eNOS has, however, been

identified in CNS neurons and PNS sensory neuron perikarya [263,667]. Accu-

mulations of NOS within axonal endbulbs act as a local enzymatic source of

nitric oxide (NO), a highly diffusible free radical that can easily spread beyond

Figure 5.1 Illustration of the sequence of events during early regeneration from

a nerve trunk transection injury. The figures from top to bottom show transection,

followed by axonal endbulb formation, early regenerative sprouts, and bridging

of the transection zone. (Illustration by Scott Rogers.) See color plate section.


the confines of the axon. Local NO release may be associated with several

actions in injured nerves including the facilitation of ectopic pain discharges,

the dilatation of local blood vessels or the alteration of local growth cone

behavior. For example, NO is capable of collapsing growth cones [265] but, in

some instances (e.g., during development), also signals axon growth through

cGMP [611]. Overall, NO may influence several aspects of regeneration depending

on the timing and extent of its release. In a preliminary study, systemic

administration of a broad spectrum NOS inhibitor improved indices of nerve

regeneration in mice [803].

Figure 5.2 Images of axonal endbulbs using immunohistochemistry to label

their specific constituents. The panels on the left show discrete (arrow) and

irregular (arrowhead) endbulbs proximal to a chronic constriction injury lesion

that contain the mu opioid receptor, double labeled with an antibody directed

against neurofilament (Nf). The panels on the right show CGRP containing

endbulbs proximal to a sciatic nerve transection also double labeled

with neurofilament. (Images reproduced with permission from [694] and

from [788].) See color plate section.


In summary, axonal endbulbs and their constituents can alter the nerve trunk

microenvironment within days of an injury. The early axon–microenvironment

interactions associated with endbulbs may in turn influence cardinal properties of

nerve repair: microvascular control, local cellular proliferation, and the induction

of pain.

Sprouting

Early axon growth cone sprouting can occur within hours after injury,

even when the axons are no longer connected to their cell body. Membrane

protrusions from neurons form the earliest outgrowths. The term “neurite”

refers to early outgrowths that are either axons or dendrites, indistinguishable

at this stage. While the full repertoire of molecules involved in early neurite

formation remains to be defined, one proteinmolecule that promotes directional

neurite outgrowth is protrudin. By interacting with other membrane proteins

such as Rab 11, protrudin helps to determine the direction of membrane sorting

and trafficking [620].

How axons acquire new lipids to manufacture membrane is important in

determining the rate of outgrowth (see review [705]). Lipids may be transported

in vesicles by anterograde fast axonal transport (microtubule dependent,

400mm/day) or may be transported on cholesterol- and spingolipid-rich lipid

rafts. Not all constituents of the outgrowing axon and growth cone need to be

manufactured at the cell body. For instance, phosphatidylcholine, the major

membrane phospholipid is locally synthesized in axons in a process that is

essential for axon outgrowth [705]. Lipid may be added all along the length of

the neuron from cell body, axon to the growth cone. In contrast, cholesterol

synthesis may be confined to the perikarya, or cell body. Apolipoprotein E may

facilitate lipid delivery [705].

After an injury in vivo, 1 to 2 days is required before axon extension from

an injury site is detected. This form of outgrowth does not occur without a

connection to the cell body [586]. Most sprouts from myelinated axons originate

within the retracted first node of Ranvier, and some from the second [195].

Friede and Bischhausen performed a detailed subserial ultrastructural analysis

of five myelinated axons transected 72 hours earlier. Sprouts were detected

within the first day. Three of five axons had extensive sprouting, whereas

the remaining two had large axonal endbulbs without sprouts (Figure 5.3).

This uneven response to injury is an important finding. Additional findings

included an orderly microtubular proliferation (tenfold rise in numbers

while retaining their axial orientation) in the proximal axon, and rises in the

density of axoplasmic organelles such as mitochondria and smooth endoplasmic

reticulum.


Specific regenerative responses to injuries

Crush injuries regenerate much more successfully than transection

because new axon sprouts can easily regenerate into basement membrane tubes

beyond the injury site (residual of previous fibers) (Figure 5.4). These tubes are

associated with proliferating SCs in the distal stump and are otherwise known as

Bands of Bungner [47,153,425]. Alternatively, transection requires that sprouting

axons make their way across a gap (the nerve is under tension and a gap results

from the injury) to find an appropriate landing site on the distal nerve stump.

As a common clinical problem, transection is a severe nerve injury that poses

major limitations on full nerve regrowth. The severe hesitancy or restraint

exhibited by early ingrowth of axons into regenerative bridges across nerve

transections is in itself remarkable. Very few axons, in comparison with the

large population residing in the proximal nerve stump, seek to enter initial

(a) (b) (c) (d) (e)

Figure 5.3 Responses of axons proximal to a rat transection injury (72h earlier)

site based on detailed ultrastructural observations by Friede and Bischhausen.

The parent axon is highlighted in black. Note the variation in growth patterns

with some axons forming multiple sprouts at either distal sites or from proximal

nodes of Ranvier. The two fibers on the right have formed very few sprouts

but demonstrate prominent endbulb formation. (Reproduced with permission

from [195].)


connective tissue bridges or cables that link proximal and distal stumps.

The explanation for the meager outgrowing axon population in this scenario is

unknown. Certainly, a suprising and substantial hostility to peripheral axon

regeneration exists.

When a peripheral nerve transection injury has been directly or primarily

sutured by a surgeon, regenerating axons do not automatically identify their

appropriate and previous pathway. Microscopic re-anastamosis for nerves by

surgeons regrettably has not solved this problem. Witzel and colleagues [744]

describe an “evolving menu of molecular cues upon which to base pathway

decisions.” Their work describes how such axons fare when they express a

green fluorescent protein (thy1-YFP) to highlight their regeneration trajectories.

Axons must decide which distal stump Bands of Bungner to enter. They arbor-

ize, travel laterally across the stump face and can have a choice of up to

100 Bands (for a single regenerating axon) to choose from (Figure 5.5). Thus,

local, rather than distal target cues determine their fate and axons make

“wrong-way” decisions often to inappropriate fascicles and targets. Some axons

grow retrogradely. Other axons can be observed to undergo Wallerian-like

degeneration, illustrating the process of “pruning.” Not all axons emerge from

the proximal stump at the same time. A small subset of early axons begin

to enter the distal stump, later followed by others, indicating “staggered”

regrowth.

Figure 5.4 Illustration of early axon sprouting following a nerve trunk crush

injury. Note that each parent axon can send multiple daughter sprouts.



The growth of axons after crush injury is usually given as 1–3mm/day, a value

linked to the rate of slow axoplasmic transport [200]. New axon sprouts emerge

from the proximal border between viable axon and degenerating axon, often, as

mentioned, at a node of Ranvier [195,572,610]. After nerve transections, rates of

regeneration are much slower. For example, after nerve transection and primary

repair with sutures, the fastest axons grew at rates of only 0.84mm/day [744].

Regeneration, however, is uneven and the average rate of all axon growth after

these more severe injuries is yet much slower than the rates given above.

As gleaned from the above discussion then, a number of important criteria

determine whether regrowth will match ideal levels. Many additional barriers

exist during later regrowth.

Staggered outgrowth

One might predict that regrowth of thousands of new axons within a

peripheral nerve trunk would take place in a uniform style when they are

injured simultaneously. This is not the case. “Staggered” outgrowth refers to

the property of axons to grow with different onsets and rates. This property

was first described in motor axons, regrowing through a zone of transection

and resuture [7]. It has also been observed when axons enter into a regenerative

bridge connecting transected nerves. Ingrowth is slow and hesitant, with only a

small proportion of axons entering the bridge at any given time. In the nerve

trunk distal to the injury site, regenerating axons can illustrate great variability

in their maturity because of their nonsychronized regrowth. Axons of different

levels of maturation, some newly myelinating, others unmyelinated are observed

[450] (Figure 5.6). Why “staggered” outgrowth occurs is uncertain. It has been

suggested that “pioneer” axons lead the way through zones devoid of obvious

guidance posts encouraging new axons to follow.

Figure 5.5 Images of axonal sprouts after a crush injury of the rat sural nerve

using immunohistochemistry to label CGRP. CGRP is displayed prominently in

early regrowing axons. (Image taken by X.-Q. Li, Zochodne laboratory.)



Growth cones and peripheral neurons

An overview of growth cones

Growth cones are the exploratory hands of axons of the regenerating

peripheral nervous system (Figures 5.7, 5.8). Their properties are not the sole

determinants of regeneration in vivo, but do have a major influence on how

subsequent regeneration will fare. In complex tissues, growth cones continually

sample their environment through extension of filopodia, retraction, and

turning. Detailed considerations of growth cone neurobiology are found else-

where [228]. In this section, we will briefly cover some relevant pathways and

molecular players. Classical growth cones consist of motile finger-like filopodia

largely built around actin filaments that emerge from web-like lamellipodia

(a) (b)

Figure 5.6 Electron micrographs of regenerative bridges traversing a conduit after

a rat sciatic nerve transection injury. The micrographs illustrate wide variation in

the maturity of the outgrowing axons because of “staggered” regeneration from

the proximal nerve stump. On the left panel, the arrow points to a cluster of

unmyelinated axons surrounded by SC cytoplasm and basement membrane.

Contrast this appearance of closely packed axons with the Remak bundle from

Figure 2.8. The arrowhead on the left shows a thinly myelinated regenerating

axon sharing its SC processes with at least two umyelinated axons, described

as a Type I regenerating unit by Morris et al. (see text). On the right panel, the short

arrow points to a regenerative cluster containing two small myelinated axons and

a number of unmyelinated axons. (Images taken by D. McDonald and reproduced

in part with permission from [450].)

95Growth cones and peripheral neurons

extending from a central area or “palm” of a “hand” where microtubules and

axoplasm terminate. The central domain (C) of the growth cone contains

splayed endings of microtubules that probe distally into what are called the

peripheral transition areas. In these zones, myosin motors are involved in the

dynamics, assembly, and depolymerization of microtubules, thereby influencing

Filopodia

Lamellipodia

ActinMicrotubules

Figure 5.7 Illustration of axon sprouts and a growth cone with its main

components labeled. (Illustration by Scott Rogers.) See color plate section.

Growthconecore

Growthconecore

Filopodia

Filopodia

Lamellipodia

Lamellipodia

Axon

Axon

Figure 5.8 Images of growth cones from adult sensory neurons harvested

and grown in vitro. (Images taken by C. Webber, Zochodne laboratory.)


growth cone behavior. The transitional zone thus found at the peripheral edge

of the C domain acts as a dynamic interface between microtubules and

actins that form filamentous (F actin) polymers. The most distal, peripheral

domain (P) of the growth cone with its web-like lamellipodia and finger-like

filopodia is the motile exploring part of the growth cone. Its properties center

on F-actin assembly and disassembly. The overall size and shape of growth cones

are therefore highly variable, thought to be larger and more complex during

pauses in outgrowth and smaller and streamlined during rapid forward growth

[54,94,98,221,442,688]. Temporally and spatially restricted calcium transients

help to determine how a growth cone turns and moves [51]. These, in turn, signal

calcium-dependent enzymes such as calmodulin dependent protein kinase II

(CaMKII; see below). A key central signaling mechanism involves the phosphati-

dylinositol-3 kinase (PI3K) pathway, in turn capable of regulating growth cone

behavior. PI3K signaling is essential to axon branch formation and axon turning

[206,473]. Molecules involved in growth cone advance or retraction and signaled

by PI3K include the RHO family GTPases discussed below, and actin and microtu-

bule interacting proteins (see review [779]). Extrinsic growth factors promote

outgrowth (advance), while other signals cause growth cone retraction (stop, or

collapse, and retreat) depending on the specific signaling cascades generated

[506]. In addition to variations in forward growth, growth cones also turn toward

(attraction) or away from (repulsion) chemotactic cues, such as gradients of

neurotrophins or netrin (see below).

Actin

Actin is assembled into filaments (F-actin) from monomers (G-actin) at

the leading tips of filopodia (þ, or barbed end), a process that requires nucle-

ation, polymerization, and annealing. The polymerizing, þ, end is associated

with ATP binding. Next, retrograde translocation of F-actin to the C domain by

myosin motors occurs, a process that helps to propel the growth cone forward.

Finally, there is depolymerization (– end or proximal part of the actin filament

bundle) in the transitional zone that is essential to recycling of actin. At the –,

depolymerized end, ATP associated with actin becomes dephosphorylated to ADP

(see review by [133]). The entire F-actin cytoskeleton is thought to turn over

rapidly, and there are changes in actin localization detectable within minutes

[226]. A series of proteins that interact with, and regulate, actin dynamics,

essential for growth cone behavior, is discussed below. In general, the assembly

and disassembly rates of F-actin are criticial for growth cone stability and

outgrowth or collapse, respectively. These rates also influence turning: asymmet-

ric F-actin assembly or disassembly in the growth cone result in changes of

direction. The motile exploratory behavior of growth cones thus depends on


polymerization and depolymerization of F-actin filaments in filopodia with

concurrent plasticity of microtubules inserted into the transitional domain of

the growth cone, discussed next [207].

Microtubules

Microtubule plasticity has a major influence on growth cone dynamics

and behavior. Microtubules form parallel networks within intact axons but

splay apart as they reach the central domain of the growth cone, sometimes

forming loops [133]. Most microtubules only rarely extend more distally. Like

actin, there are distal þ ends where tubulin dimers are added: a-tubulin and

b-tubulin. Depolymerization of the mictotubule occurs at the – end. The poly-

merizing zone or þ end has dynamic instability with properties that include

shrinkage or “catastrophe,” and growth, termed “rescue.” These properties are

critical for growth cone exploratory behavior and they are regulated by a series

of proteins known as MAPs (microtubule associated proteins) or MBPs (micro-

tubule binding proteins). “Structural” MAPs that stabilize microtubules include

MAP1B, MAP2, and tau (the latter is important in CNS neurodegenerative dis-

orders), whereas others act as motors for axoplasmic transport (dyneins, kine-

sins). Further MAPs are called þ end tracking proteins (þTIPs) that are thought

to influence microtubule dynamics (see [133] for an extensive list of MAPs).

Analogous to ATP-actin polymerization, b-tubulin is also associated with GTP

during polymerization. As the microtubule polymer “ages,” GTP-b-tubulin is

then converted to GDP-b-tubulin prior to depolymerization. “Aging” occurs

from the þ end toward the – end of the microtubule, or distal to proximal.

During subsequent depolymerization, b-tubulin and a-tubulin dimers are thus

released at the – end of the microtubule. a-tubulin is also tyrosinated at the þend of the microtubule during polymerization and detyrosinated at the – end

during depolymerization [133]. Acetylation of tubulin also influences growth

cone stability.

Microtubule severing, like that of actin above, has an essential role in axon

outgrowth. Katanin is a heterodimeric protein severing enzyme that creates

multiple short microtubules and allows greater microtubule plasticity. A fine

balance between excessive severing and inadequate severing is critical in

optimizing microtubular dynamics [255,322].

Molecules that influence growth cone behavior

Growth cone receptors include Trks (tropomyosin-related kinase; high

affinity neurotrophin receptors) that are selective for specific neurotrophins

(e.g., NGF for TrkA, BDNF for TrkB) but also p75, expressed widely on axons


(formerly called the “low affinity nerve growth factor receptor”). Recent work has

also suggested that proneurotrophin molecules, precursor molecules of classical

neurotrophins, specifically bind p75 receptors and induce widespread actions on

axons [293,383]. p75 activation alters the behavior of the Ras superfamily GTPases.

These molecules operate as molecular switches in altering cell polarity and

symmetry and are localized in growth cones [175]. Overall, GTPase behavior

involves translocation to the growth cone membrane, a property that can be

identified using newer imaging techniques. Three members, RHOA, RAC1, and

CDC42 have substantial but differing influences on neuronal growth cone dy-

namics [217,769]. Therefore p75, when occupied by a ligand (e.g., a proneurotro-

phin), inhibits RHOA activation, an action that facilitates outgrowth

[175,213,217,758]. In contrast, when unoccupied or unligated, p75 may instead

activate RHOA [213] to promote growth cone collapse. This is one of many

examples of plastic function among growth-related molecules, depending on

the context and type of interaction.

RHOA, the most extensively studied of the Ras GTPases in growth cones, is a

growth cone brake. It acts through RHO kinase (ROK) that enhances myosin II

phosphorylation and subsequent actin mediated growth cone retraction

[217,506]. RHOA may accomplish retraction in other ways as well, such as inhibit-

ing of myosin phosphatase activity; this action leads to increased myosin ATPase

activity and growth cone retraction [429]. The net result of ROK activation

responsible for retraction are increases in actin “arc” formation and central

actin bundle contractility and stability. RHOA localizes not only to growth cones

but also to the leading edges of migrating cells where it can promote its actions

[553] (Figure 5.9).

In the CNS, RHOA activation is associated with impaired regeneration of

spinal cord axons [122,134]. Specific CNS mediators of RHOA-mediated regenera-

tive failure include the myelin-associated molecule Nogo-66, myelin-associated

glycoprotein, semaphorins, ephrins, oligodendrocyte myelin glycoprotein,

and chondroitin sulphate proteoglycans [5,44,52,185,188,302]. Most of these

molecules generate complex interactions with the Nogo receptor (NgR),

p75, LINGO-1, TAJ/TROY, calcium, and the epidermal growth factor receptor

(EGFR) [183,468,615], and eventually RHOA. In the PNS, SC-derived growth factors

are capable of disrupting these kinds of interactions through regulated intra-

membranous proteolysis of molecules involved in the cascade, extracellular

interruption of NgR binding, and intracellular disruption of NgR p75 inter-

actions [5,6]. Thus, they interrupt RHOA activation. mDia is an additional down-

stream RHO effector thought to influence axon lengthening [48]. Transfection

with a adeno-associated virus containing the inhibitor C3 ADP-ribosyltransferase

that inactivates RHOA in retinal ganglion cells [185], or use of a cell permeable


modified version of C3 ADP-ribosyltransferase, increased regeneration in CNS

neurons [44,185].

RHO GTPases play key roles in growth cone turning, a determinant of whether

misdirected axons compromise regeneration [217,769]. Turning behavior is

dependent on asymmetric release of intracellular calcium within the growth

cone [311]. In keeping with this mechanism, RHOA levels decline in growth cones

that are undergoing attractive turning in response to a calcium-mediated signal

generated by ligation of BDNF or netrin. This allows one portion of the growth

cone to expand. In other areas of the growth cone, concurrent RHOA upregula-

tion may mediate active repulsion [311]. Thus, rises and falls in local RHOA

molecules associated with the growth cone membrane determine whether it

actively promotes repulsion or allows attraction, respectively.

Figure 5.9 Simplified drawing illustrating how growth cones change the direction

of their trajectory in response to localized gradients of growth factors such as

NGF acting on their TrkA receptors (green). RHOA GTPase (red) is thought to mediate

active repulsion and may be localized to growth cone membranes that are collapsing

and not advancing whereas RAC1 facilitates outgrowth and may be localized to the

membranes of advancing parts of the growth cone. See color plate section.


RHO and ROK are mediators of actin regulation. RHOA-ROK activates LIM kinase,

a pathway that in turn phosphorylates cofilin, also known as actin depolymer-

izing factor (ADF). When cofilin is phosphorylated, it fails to bind actin and to

participate in normal actin turnover. Altered phosphorylated cofilin function

therefore interferes with growth cone advance [214,505]. To exert its effects

“normally”, active nonphosphorylated cofilin binds to filamentous actin (F-actin)

that is bound in turn to ADP. The proximal end (–) of the actin filament in

lamellopodia undergoes depolymerization when bound. The units of cofilin then

disassociate from released units of actin, now in the form of G-actin (globular

depolymerized). Actin-interacting protein 1 works with cofilin to disassemble

actin filaments by capping actin filaments that have been severed. Capped

actin is not available to reassemble and depolymerization is thereby promoted.

“Slingshot” is yet another protein that acts as a cofilin phosphatase, dephospho-

sphorylating it (activating it) and thereby enhancing actin turnover [168,507,723].

Further proteins that inhibit actin binding include phosphatidylinositol

4,5-bis-phosphate (PI(2)P) and tropomyosin.

Cofilin has complex roles. We have just discussed its activity during actin

depolymerization, but it is also capable of promoting active actin polymeriza-

tion. At the þ barbed (forward) assembly end of the actin filament, G-actin

interacts with cofilin and ATP to begin polymerization. Overall, active cofilin

therefore has a number of roles that induce cell protrusion and set the direction

of cell migration [214]. How the complete repertoire of proteins (including those

that interact with cofilin) is critical to growth cone dynamics has only been

recently clarified (for review, see Ono [536]).

The proteins that regulate actin filament dynamics may have considerable

influence on the fate of peripheral nerve regeneration. A more complete list of

their roles can be found elsewhere (see reviews [133,465]). Many are localized

with F-actin at the leading edges of growth cones [226]. Arp2/3 is a complex of

seven proteins (Arp2, Arp3, p41Arc, p34Arc, p21 Arc, p20Arc, and p16Arc) that

nucleates new actin branches from existing actin for outgrowth of the growth

cone. N-WASP, cortactin, and Scar/WAVE are activators of ARP2/3. Ena/Mena/VASP

proteins block capping of the barbed (þ) ends of actin, thereby allowing longer

filopodia to form [226]. Barbed end actin capping proteins include Cap Z and

actin monomer binding proteins (thymosin b4 and profilin) that are expected

to reduce growth cone advance [34,493,702]. Profilin, however, also sequesters

G-actin monomers to facilitate actin polymerization, thus enhancing filopodia

formation. Proteins of the Ena/Mena/VASP family also enhance polymerization

rates by recruiting profilin to sites of actin assembly [226]. Thus, profilin is

another example of a molecule with apparent roles in both growth cone advance

and collapse [343,761].


CDC42 and RAC, also members of the Ras GTPase family, facilitate growth cone

advance. Their behavior thus differs substantially from that of RHOA. CDC42

promotes filopodia and lamellipodia extension. Rac influences actin filament

behavior, possibly by interacting with cofilin. It also stabilizes adherent contacts,

an important step in promoting outgrowth [217,429]. RAC1 signals through the

kinase PAK (p21-activated kinase) that activates NFkB and MAPK and increases

neurite outgrowth [127,196,197].

Despite all of this critical information, the significance of RHO family GTPase

function in the peripheral nervous system is largely unexplored. Peripheral

neurons can be coaxed in vitro to grow across inhibitory CNS substrates by

inhibiting RHO pathways [5,52,302]. Wu et al. [749] recently demonstrated

that embryonic dorsal root ganglion neurons had RHOA synthesis localized

to developing axons and growth cones. At this site, semaphorin 3A, a molecule

that repels axons by collapsing growth cones (see below under GSK3b), signaled

growth cone axons to synthesize RHOA protein. Despite these findings, however,

we do not know how adult peripheral axons might be influenced by RHOA

in peripheral nerve trunks. RHOA is expressed and functional in the peripheral

nervous system, and there is preliminary evidence that it has actions in the

PNS similar to those in the CNS. For example, Terashima and colleagues [673]

identified RHOA, RAC1, CDC42, and PAK in intact peripheral nerve trunks and

ganglia. They were localized to sensory neuron plasmalemmae but were also

localized in axons and SCs further out in the nerve trunk. Hiraga and colleagues

[270] noted that fasudil (HA-1077), a pharmacological ROK antagonist given

systemically in mice, increased the recovery of motor reinnervation (CMAPs)

and the caliber of their regenerating axons. Activated RHOA was identified in

the portion of the spinal cord where motor neurons reside, but not in distal

stumps of the injured nerves. In separate work, fasudil increased the neurite

outgrowth of adult sensory neurons even when grown on noninhibitory sub-

strates in vitro. It also increased the growth of axons through regeneration

conduits spanning nerve transections in rats [105]. There are important inhibi-

tory cues in peripheral nerves that may activate RHOA such as myelin-associated

glycoprotein (MAG) and chondroitin sulphate proteoglycans.

RHOA-ROK and other Ras GTPase family members also influence microtubular

organization [774]. RAC captures and stabilizes microtubules through molecules

called IQGAP1, CLASP, and Clip-170 [779]. CDC42 acting through the protein

Par3/6-aPKC inactivates GSK3b, discussed below, and thereby promotes microtu-

bule assembly through a molecule called APC (adenomatous polyposis coli tumor

suppressor protein) [779].

cAMP is a critical mediator and facilitator of growth cone behavior, especially

during guidance [412,474,636]. By increasing the content of cAMP, or “priming”


in sensory neurons, for instance, their outgrowth inhibition to MAG can

be overcome [84,564]. Interestingly, cAMP also facilitates how SCs support out-

growth, discussed below.

Glycogen synthase kinase 3b (GSK3b) is multifunctional serine/threonine

kinase, localized to the leading edges of growth cones. In its constitutively

active nonphosphorylated state, it suppresses growth cone extension and axon

formation [165]. When it is phosphorylated through the PI3K-Akt (PKB) signal-

ing pathway, GSK3b activity is shut down, allowing growth to occur. This

pathway was originally linked to neuroprotection and survival, providing one

further example of how axon outgrowth and neuronal survival signaling

pathways are closely linked [113,143]. GSK3b inhibits axon outgrowth in part

by targeting APC mentioned above [780]. GSK3b can also be inactivated by NGF

through PI3K, by dishevelled proteins of the Wnt signaling pathway, and also

by preconditioning of neurons subsequently grown on laminin to facilitate

integrin binding [779]. Integrins are the receptors on cells for extracellular

basement membrane components. As suggested, basement membrane–integrin

interactions also involve signaling properties beyond simple adhesion (see

Chapter 10) [780].

PI3K (phosphatidylinositol 3-kinase)-Akt [70] is a critical downstream survival

pathway that mediates trophic factor support of neurons and blocks apoptosis.

Loss of neurons through apoptosis is an important consideration during regen-

eration. For example, retrograde apoptotic loss of neurons after axotomy, more

prominent in neonates, reduces the pool of parent neurons available to send

out regenerative sprouts. Within individual neurons, however, it may be that

PI3K-Akt activation serves roles that promote growth and differentiation (for

review see [189]). PI3K increases the conversion of PI-4,5-P2 (PI(2)P; phosphatidy-

linositol 4,5-bis-phosphate) to PI-3,4,5-P3 [PI(3)P] that in turn phosphorylates and

activates the downstream molecule Akt (also known as PKB). PTEN (phosphatase

and tensin homolog deleted on chromosome ten) is a phosphatase that inhibits

the conversion to PI(3)P. Predictably, molecules that suppress PI3K activity

such as PTEN facilitate growth cone collapse [95]. Like GSK3b, PTEN is also

inactivated by phosphorylation, including one pathway involving NGF acting

on p75 receptors independently of TrkA. A second pathway involves a kinase

known as CK2 (casein kinase II) [22] that also phosphorylates and inactivates

PTEN. Alternatively, growth inhibitor molecules may activate PTEN. Thus,

several pathways can converge on PTEN to block it and promote axon

outgrowth.

CRMPs (collapsin-response mediator proteins) are also involved in growth cone

collapse. This family, including members CRMP 1 through 5, are cytosolic phos-

phoproteins that influence the formation of microtubules through tubulin


binding [229]. When CRMP 2 is phosphorylated by active GSK3b, axon formation

is inhibited (for review see Li[399]). CRMPs interact with semaphorins, a family of

proteins both secreted and membrane bound, that influence axon guidance.

Through their neuropilin receptors (NP-1 and NP-2) and their interactions with

CRMPs, they are also critical mediators of growth cone collapse [544,596]. Class 3

semaphorins, or Sema3s are secreted repulsive molecules and among them

Sema3A has had the most attention. Semaphorin-mediated growth cone collapse

has been specifically linked to CRMP2, in turn inhibiting microtubule assembly

[201,602]. There are further series of associated protein interactions that can be

initiated by Sema3 to collapse growth cones. These include not only phosphoryl-

ated CRMP 2, but also a2-chimaerin, CDK-5/p35, and RAC [67]. The participation

of RAC in the collapse of growth cones is surprising, considering our discussion

above of its facilitatory role. RAC may, in fact, shuttle between dual roles

promoting either outgrowth or collapse depending on the multimolecular

scaffold it interacts with. CDK-5 may also phosphorylate CRMP 2 together with

GSK3b [120]. Sema3A operates through PTEN to reduce PI(3)P levels, reduce Akt

signaling, and to prevent GSK3b inactivation. These steps lead to growth cone

collapse. To allow these interactions to occur, close localization is important.

Thus, both Sema3A and PTEN are localized to growth cone membranes [95]. These

interactions assume relevance in peripheral nerve regeneration because Sema3A

is expressed during Wallerian degeneration.

Sema3A is expressed in cranial and spinal motor neurons, and its levels in

these neurons decline after peripheral axotomy [544]. It is also expressed in

terminal SCs of neuromuscular junctions (see Chapter 6) and may help influence

the reinnervation of specific fiber types. Moreover, other cell types express

Sema3s, and its presence in the extracellular matrix allows growth cones to be

exposed to gradients of the molecule [544]. All of these forms of expression

indicate that Semas play a central role in both ligand signaling to growth cones,

and also in subsequent signal transduction within neurons.

CRMP4, in particular CRMP4b, binds to RHOA within growth cones to inhibit

neurite outgrowth [8,585]. Knockdown of CRMP4b or binding competition using

a portion of its N terminus (known as C4RIP, or CRMP4b-RHOA inhibitory peptide)

allows sensory neurons to grow on inhibitory substrates such as myelin or

aggrecan, a chondroitin sulphate proteoglycan.

Protein kinase C (PKC) and GAP43/B50

PKC subtypes (both calcium-dependent a, b, and g subtypes and calcium-

independent d, e, and z) are localized to regenerating growth cones in vivo,

especially their plasma membranes [329,531]. This is an important finding in


metabolically active growth cones, indicating their role as a substrate for signal

transduction, regulation of ion channel function, growth, and modulation of

local protein production. PKC modulates how GAP43/B50 (growth-associated

protein 43 or B50), another central signaling molecule, is upregulated by injury

(see below) and how it alters growth cone behavior. GAP43/B50, originally dis-

covered by Zwiers, Ostreicher, and colleagues [10,119,525,526], and later by

Skene and colleagues [630] is associated with the cytoplasmic surface of the

growth cone membranes. It alters growth cone plasticity when membrane

bound and may operate to amplify extracellular signals [296]. In a hypothesis

developed by Ostreicher, Gispen, Zwiers, and others [119,524,709], calcium

entry into axon terminals causes the release of GAP43/B50 from binding with

calmodulin. GAP43/B50 then is phosphorylated by PKC and can interact with

F-actin at growth cone membranes. It is unclear whether the phosphorylation of

GAP43/B50 influences its membrane localization, but when it undergoes ADP

ribosylation it is released from plasma membranes into the cytosol [524]. CAP-23,

a related molecule also upregulated after injury, may offer similar actions [443].

In growth cones, calmodulin that is released from its association with GAP43/B50

also has diverse actions, including membrane recycling, excytotic/endocytotic

activity, and polymerization of microtubules and actin. It also phosphorylates

GAP43/B50. A dominant inhibitor of calmodulin stalled the growth of CNS

pioneer axons [343].

More recent work has updated how PKC, GAP43/B50,CAP-23, and a related

molecule known as MARCKS (myristoylated alanine-rich C kinase substrate)

influence growth cones GAP43/B50, CAP-23, and MARCKS are expressed in

membrane raft microdomains and control PI-4,5-P2 availability at the growth

cone. PI-4,5-P2, GAP43/B50, and CAP23 are all capable of altering actin dynamics

in concert with actin binding proteins. From above, we also know that PI-4,5-P2

is converted to PI-3,4,5-P3 by phosphatidylinositol 3,4,5-triphosphate (PI(3)K) to

facilitate axon outgrowth [191,378]. As an aside, both are also expressed in SCs

[443,556]. PKC may play a role in laminin–integrin interactions, and local neuro-

trophin signaling [676]. These are aspects of regeneration mentioned earlier but

considered in more detail in later chapters. CaMKI (calmodulin kinase I) is

expressed throughout the neuron including growth cones where it acts as a

positive transducer of growth cone motility and axonal outgrowth [729]. It has

not been studied in peripheral neuron growth cones.

In summary, localized calcium transients signal calmodulin and CAMKI or II,

calpain, PKC, as well as the inhibitor phosphatase calcineurin. An eventual

outcome is a change in the phosphorylation status and localization of GAP43/

B50, features that influence its role as an actin capping factor [51].


In vivo growth cones

Few descriptions of in vivo growth cone morphology are available. They

can be difficult to accurately identify in histological preparations. Ultrastruc-

tural studies provide some descriptions [740]: enlarged axonal-like profiles with

increased cytoplasmic electron density, and accumulations of smooth endoplas-

mic reticulum. Scanning electron microscopy depicts them as enlarged (3–5mm

in diameter) axon structures that are highly variable in configuration but have

relatively few filopodia or F actin bundles. They have a close and almost invari-

able relationship to SCs and their basement membrane [364,532]. In vivo growth

cones also include heterogenous vesicles, mitochondria, and peripheral accumu-

lations of electron dense filamentous material. Neurofilaments are not usually

identified within them since the neurofilament protein polymer lattice instead

ends proximal to the site of growth cone formation. Using immunohistochem-

istry, the distal portions of axons with growth cones can be labeled with PGP 9.5,

bIII tubulin or GAP43/B50. All of these markers highlight the complexity of

in vivo growth cones [450] (Figure 5.10). Some proteins have been localized using

immuno-EM. For example, there is diffuse cytoplasmic expression of synaptophy-

sin, an important component of the synaptic vesicle membrane [532]. This

protein regulates the exocytosis of vesicles and is prominent in early regener-

ating sprouts arising from nodes of Ranvier.

Early SC behavior

In the earliest phases of regeneration, there are dramatic changes in

SC phenotype and behavior that have important impacts on axon regrowth

(Figure 5.11). By day 5–7 after a peripheral nerve injury, there is an initial wave

of Schwann cell (SC) proliferation in both the proximal and distal stumps, as well

as proliferation of other cellular constituents including mast cells, endothelial

cells (angiogenesis), and fibroblasts [234,787,805,807]. The classical view of SC

behavior suggests that SCs first alter their phenotype secondary to loss of contact

with the axon. This behavior is most prominent in those SCs associated with

myelinated axons where the integrity of the myelin sheath critically depends on

the presence of an adjacent axon. Once that contact is interrupted, SCs become

“reactive.” They dissolve their own myelin, downregulate their synthesis of

myelin protein mRNAs, and begin to proliferate, forming Bands of Bungner

as guideposts for new axons. Proteins rapidly upregulated by SCs following

interruption of axon contact include p75, the low affinity nerve growth factor

receptor discussed above (and also expressed on SCs) and GFAP (glial fibrillary

acidic protein) [106,307,475]. Transgenic mice lacking GFAP develop normally,


StableMyelinating

SCs GFAPerbB2

activationC-J un

activation

p75

DedifferentiationProliferation

Figure 5.11 Illustration of changes in SC phenotype after nerve injury.


Figure 5.10 Examples of complex in vivo growth cones found at the proximal stump

of a transected rat sciatic nerve and labeled using immunohistochemistry for

neurofilament (red) or PGP 9.5 (green). Note that the appearance is substantially

different from growth cones studied in vitro. The PGP 9.5 labels the full extent of

the structure, whereas neurofilament is only found in portions of the growth cone

and in its proximal stem. (Bar ¼ 20 microns) (Reproduced with permission from [450].)


107Early SC behavior

but lack appropriate plasticity after injury and regenerate more slowly [693].

At somewhat later time points, SCs upregulate GAP43/B50, already discussed

above in regenerating axons [126]. It is likely that there is a continuum of

SC plasticity that responds to changes in the local microenvironment. Like

the immune system, sequential and coordinated changes in behavior and

protein expression of this class of cells illustrate a surprising beauty of planning

and action.

The first signals for this new behavior of previously “quiescent” SCs are likely

to be more subtle and interesting than a simple lack of contact. Indeed, some of

the protein markers that are rapidly upregulated in stable SCs ensheathing

myelinated axons are expressed in “intact” SCs comprising the Remak bundle

of unmyelinated axons: GFAP, N-CAM, A5E3, Ran-2, and p75 [308]. This supports

the idea that Remak SCs are much more plastic and dynamic than expected [126].

While GFAP is typically expressed in Remak SCs and not in the SCs of myelinated

axons, the latter cells begin to express it after a nerve injury, even if they are

remote from the injury site [108]. A rapid signal capable of facilitating dynamic

SC behavior may be at play, possibly heralded by axon atrophy and distortion of

the inner mesaxon of the myelin sheath. As will be discussed below, the molecule

NRG (neuregulin) may be partially responsible for changing SC behavior [172]. In

their subserial ultrastructural study of transected myelinated axons, Friede and

Bishhausen [195] described asymmetric hypertrophy and additional cytoplasmic

changes of SCs near the injury site and as far proximal as two nodes of Ranvier.

These changes were found in portions of the SC found closest to the site of axon

interruption. They included microtubular profliferation, increased rough endo-

plasmic reticulum, and formation of SC processes or tongues resembling those of

Bands of Bungner. These findings indicate a remarkable alteration of properties

that increases, even within portions of a single SC cell, near an injury site in a

proximal stump.

An interesting and parallel form of behavior involves the perineuronal satel-

lite cells, cousins of SCs, that normally surround and support neurons in dorsal

root ganglia. Normally these cytoplasmic-poor cells can be barely resolved as they

intimately wrap around the perikarya of sensory neurons. After injury, they

become hypertrophic and upregulate their expression of GFAP, forming ring-like

profiles around each sensory neuron [103] (Figure 5.12).

The molecules and pathways that are important for SC proliferation partake

in a complex synergy between cAMP and growth factors. For DNA synthesis in SCs

to occur, as required for proliferation, a rise in SC cAMP levels is a necessary first

step, priming subsequent actions by NRG, PDGF, or bFGF [643]. Experimentally,

cAMP levels can be boosted by activating adenyl cyclase through pharmaco-

logical approaches. Other SC mitogens include CGRP [111], insulin, insulin-like


growth factors I and II (IGFI and II) [633], and others. GFAP, discussed above, is

also important in signaling SCs to proliferate. It does this through an interaction

with integrin avb8 involving ERK phosphorylation [693]. Parathyroid hormone-

related peptide (PTHrP), synthesized by SCs, is important in promoting

SC migration and alignment or bundling with axons and collagen fibers [432].

It may signal SCs by the phosphorylation of CREB.

There are a large number of additional molecules that have an important

influence on SC behavior. For example, p21 and p16 within SC nuclei act to

shut down their proliferation [26]. IL-6 and LIF can act to activate SCs to

produce further LIF or MCP-1, a macrophage chemoattractant [682]. When

axons contact SCs, however, SC LIF mRNA levels are turned back down [447].

The low-density lipoprotein receptor-related protein (LRP-1) is upregulated in SCs

after injury [86]. This transmembrane protein, a member of the low-density

lipoprotein receptor family, is widely expressed and is acted on by over

40 ligands. LRP-1 possesses cytoplasmic signaling motifs, some of which influ-

ence SC survival through upregulation of phosphorylated Akt. Its upregulation

in SCs may be driven by TNF-a. Thus, LRP-1 ligands, like PDGF, NT-3 and IGF-II

above, may serve as autocrine SC survival factors that are important during

regenerative dedifferentiation, proliferation, and migration [462].

Figure 5.12 Immunohistochemical images of a DRG proximal to a sciatic nerve

injury carried out 5 days earlier labeled with GFAP (glial fibrillary acidic protein).

GFAP labels glial cells including perineuronal satellite cells. These cells, closely

apposed to and surrounding sensory neurons, become hypertrophic and upregulate

their content of GFAP after an injury (arrows). (Images taken by Chu Cheng,

Zochodne laboratory.) See color plate section.

109Early SC behavior

Local axon–SC interactions and early outgrowth

The axon–SC relationship

SCs are essential partners in axon guidance and in the elaboration

of growth factors and adhesion molecules for regrowth (Figure 5.13). Thus,

SC support extends beyond simple provision of laminin basement membrane

scaffolds (Figure 5.14). As we have discussed, the SC phenotype can change

rapidly after injury, even preceding frank Wallerian degeneration, indicating

SCs that are closely “attuned” to axon events [108]. These aspects of partnership

are described next.

The support offered by SCs has been illustrated by recent work transplanting

cells into nerve transection gaps. Stem cells (e.g., from hair follicles) that

are capable of differentiating into SCs, or transplanted syngeneic SCs, when

implanted into sciatic nerve transection gaps substantially improve axonal

regeneration [13,241,455,482]. Using another approach, axon outgrowth onto

bioartificial films was studied. If reactive SCs were seeded onto the film substrate

in advance to partner with new axons, there was enhanced outgrowth [687].

Conversely, irradiation of regenerating nerve segments severely impaired nerve

regeneration by preventing SC proliferation [601]. Similarly, the inhibition of SC

mitosis using mitomycin prevented axon and SC ingrowth into acellular nerve

grafts [247]. While axons therefore depend on SCs for outgrowth, SCs may not

need axons to grow out across nerve transection gaps. Without axons, how-

ever, they do so with substantially less directional specificity. “Pure” SC out-

growths lacking axons can produce “pseudo-nerves,” macroscopic tissue bridges

that resemble normal nerve bridges [776]. Overall, these observations highlight

the essential partnership or “dance” between the axon and SC during early

regeneration.

The close, almost invariable relationship between axon and SC outgrowth

has been confirmed in several studies [103,426,450]. Proliferating and migrating

SCs send long processes that appear to lead and guide axons in an outward

direction from their transection site [103] (Figure 5.15). Most regenerating

axons follow SC processes in exquisite detail through three dimensions of space;

invariably diving and turning together. In previous studies, Ramon y Cajal [572]

and others [295] suggested that early outgrowing axons are often “naked,”

traveling alone without a SC partner. While examples of this exist, double label

immunohistochemistry, confocal microscopy, and electron microscopy have

more recently indicated that they are the exception. Immunohistochemical

markers such as bIII tubulin identify fine and irregular axon growth patterns

at their furthest limits, probably indentifying growth cones “sampling” distal


Figure

5.13

Illustrationsh

owingthesequen

ceofch

anges

insensory

neu

ronsandSCsafter

atransectioninjury.Note

thedevelopmen

tof

hyp

ertrophic

perineu

ronalsatellitecells.Neu

ronshave

peripheraldisplacemen

toftheirNissl

substance

andtheirnuclei.Invasionof

hem

atogen

ousmacrophages

accompaniesproliferationandded

ifferentiationofSCspreparingthewayfornew

axonoutgrowth.(Illustrationby

Scott

Rogers.)See

colorplate

section.

111

terrain. Most growth cones, however, appear to routinely accompany SC pro-

cesses. Their behavior replicates developmental growth where Schwann cell

precursors are close and consistent cellular companions of growth cones as they

approach targets [725]. During development, one analysis suggested that SC

precursor cytoplasmic processes covered over 80% of the growth cone.

Regenerative units

Morris and colleagues [478] describe two types of regenerative axon-SC

units that begin to emerge from the proximal stumps of transected peripheral

nerves: Type I units consist of an apparently mature myelinated axon sharing a

single SC with one or more unmyelinated axons. The unmyelinated axons are

applied to the SC surface beneath the basement membrane (BM) and the unit

thereby forms one axon–SC–BM complex. Type II units consist of nonmyelinated

axons only. These axons, unlike normal Remak bundles, grow in clusters and are

not always separated by SC cytoplasmic leaves. Complex and sometimes redupli-

cated BM surrounds the units. Axon sizes are also more variable in a regenerative

unit than in a normal Remak bundle and a single SC might be associated with

Figure 5.14 Adult rat peripheral sensory neurons in vitro showing extensive neurite

formation. The neurons underwent a preharvesting preconditioning injury and are

examined at day 3. The arrow points to a collection of glial cells in the culture that

have attracted neurites to grow towards it. The image illustrates the tropic properties

of supporting cells during axon outgrowth. (Culture prepared by Sophie Dong,

Zochodne laboratory.) See color plate section.


up to 80 axons. Both Type I and II units probably arise from single parent

axons. Overall, regenerating units are also associated with “halos” of small

caliber (30nM) collagen fibrils. Regenerative units are illustrated in Figure 5.16.

In the original ultrastructural descriptions, some additional features of

regenerating units were identified. For example, with time, one or more axons

in Type II units became myelinated. Near the injury zone, Morris and colleagues

[477,478] reported that frequent SCs contained membranous debris from phago-

cytosis, while at the same time associating with axons. This is an interesting

finding indicating that individual SCs can have multiple concurrent roles that

include phagocytosis while actively supporting new regenerating axons.

Other morphological changes of “reactive” SCs included alterations in their

endoplasmic reticulum and even the presence of cilia.

Figure 5.15 Immunohistochemical images of axon (labeled with an antibody

directed against neurofilament) and SC (labeled with an antibody to GFAP) outgrowth

at day 7 from the proximal stumps of transected rat sciatic nerves. The images in the

top panel illustrate extensive outgrowth close to the proximal stump and the images

in the bottom panel are further distally. Note the close relationship between axons

and SCs. (Images taken by Y. Y. Chen, Zochodne laboratory.) See color plate section.

113Local axon–SC interactions and early outgrowth

Figure 5.16 Examples of early myelinated axon repopulation of regenerative bridges

growing through a conduit connecting the proximal and distal stump of a transected

rat sciatic nerve at day 21. Note that the axons are often in clumps resembling

minifascicles. The inset shows a higher power view of two small clusters of myelinated

and unmyelinated axons. The images are from semithin sections, epon embedded and

stained with toluidine blue. (Bar ¼ 20 microns)


Local axon–SC interactions during early outgrowth can be further studied

using a regeneration conduit connecting the proximal and distal stumps of a

severed nerve. Conduits allow an analysis of early axon and SC outgrowth that is

free from adjacent products of Wallerian degeneration found in the distal stump

proper. Surgical literature on the use of conduits, largely emphasizing variations

in approach or design that might enhance regeneration, is available. Their use

has become widespread. Biodegradable conduits have been used surgically

[320,423,428,775] while experimentally they have been used to instill growth

factors [320], genetically engineered SCs or SC precursors. Immunohistochemis-

try to label axons (neurofilament, bIII tubulin, PGP9.5), SCs (GFAP), and other

constituents can be examined during early regrowth after transection and

conduit placement. For example, within the first week following transection,

the approach can evaluate the number of outgrowing axons, their directionality,

their association with SCs, and the furthest distance they have regrown. Regen-

erative bridges can be examined for the caliber (maturity), myelination, and

number of axons by LM or EM somewhat later (e.g., 2–3 weeks). If each stump

is placed in a conduit, a coaxial fibrin matrix cable connects the proximal and

distal stump by 1 week. By 2 weeks, there is migration of Schwann cells, fibro-

blasts, and endothelial vascular cells from both stumps, all of which precede

later ingrowth of axons from the proximal stump. The initial ingrowing axons

appear in clusters. They are unmyelinated, and similar to the Type II regener-

ating units described above. As axons mature, they are joined by other new

axons, collectively forming compartments or minifascicles.

Directionality

Complex patterns of regrowth are typically a consequence of peripheral

nerve injury [740]. How do axons know which direction to grow? Classical experi-

ments by Lundborg and colleagues demonstrate that axons prefer to grow

toward their own distal stump [426,492]. Transection injuries place much greater

demands on direction finding from peripheral nerve axons of the proximal

stump. After nerve transection, there is retraction of both the distal and prox-

imal nerve away from the transection site (e.g., by 3–5mm in the mid-sciatic

nerve of rats) because the nerve is usually under some degree of tension. The upper

limit for successful growth across nerve gaps in rats generally is 10–15mm.

If placed in a conduit connecting the proximal and distal stump, axons from

the proximal stump forge ahead more successfully if a distal nerve stump

instead of a portion of unrelated tissue is placed at the far end of the conduit.

Thus there is an important tropic effect of distal nerve tissue on regrowth

of the nerve. In many instances, both SC and axon also turn together but


incorrectly; they can even extend backwards into the epineurium of the prox-

imal nerve stump. We have called these “wrong way axons” (Figures 5.17, 5.18).

SCs whose proliferation was arrested, for example, by exposure to local mitomy-

cin, a mitosis inhibitor, fail to properly support regrowth. Axon regeneration was

severely attenuated and misdirected [103].

Laminin and other components of SCs

Leading SCs may elaborate trails of laminin for detailed guidance of

following axons. Laminin receptors on axons, known as integrins (see Chapter 10),

may then allow axons to adhere to these trails and closely follow trajectories of

SCs. It may also be that laminin from more than one source could help axons

to navigate. For example, vascular laminin is elaborated and is associated with

ingrowth of endothelial cells into new nerve trunks. Axons growing into regen-

erative bridges in conduits, however, did not appear to track vascular laminin.

Nonetheless, simple infusion, of soluble laminin into regeneration conduits

improved axon growth, directionality, and repopulation [103] by being incorpor-

ated into the regenerative bridge. Laminin infusion, therefore, can partly com-

pensate for a failure of SC support. This is not surprising since in vitro growth of

neonatal sensory axons is facilitated by laminin-coated filaments in the absence

of SCs [573]. Thus, laminin, elaborated by SCs may be a key link that confers

directional specificity and outgrowth enhancement to following axons.

In addition to laminin and other basement membrane extracellular mol-

ecules (including tenascin C, fibronectin, and heparan sulphate proteoglycan),

Figure 5.17 Immunohistochemical images of axons emerging from the proximal

stump of a transected rat sciatic nerve (connected by a conduit) at day 7 labeled

with an antibody to neurofilament. Note the rapid fall off in axon numbers as

they penetrate the regenerative bridge. The inset and arrow shows an axon in a

misdirected, or “wrong way” trajectory. (Reproduced with permission from [450].)



(a)

(b)

(c)

Figure 5.18 Immunohistochemical images of fine axons at the most distal end of

regenerative outgrowth by day 7 following transection of a rat sciatic nerve

(connected by a conduit). The axons are labeled by antibodies to bIII tubulin and

neurofilament. The small arrow shows a double labeled profile. The arrowhead and

large arrow show a tubulin, but not neurofilament labeled axon profile that is

highly irregular and partly misdirected. The images illustrate that very fine distal

axon outgrowth at the regenerative front may not label with a neurofilament

antibody. (Bar ¼ 50 microns) (Image taken by Chu Cheng, Zochodne laboratory

and reproduced with permission from [450].) See color plate section.


SCs also synthesize additional molecules likely to influence peripheral nerve

regeneration [200]. These molecules can support either the SCs themselves,

known as autocrine support, or axon partners. Specific examples include the

classical neurotrophins (NGF, BDNF, NT-3, NT-4/5), proneurotrophin precursors,

neuropoietic cytokines (including CNTF, IL-6, oncostatin, and LIF), other neuro-

trophic factors (including insulin, IGF-1, and GDNF), fibroblast growth factors

(FGFs), and cell adhesion molecules (including NCAM, L1, and N-cadherin). Exam-

ined in more detail in Chapter 9, growth factors in particular can signal both

locally and centrally following retrograde transport, and they appear capable of

supporting both neuron survival and axon outgrowth [100,466,489,566]. One of

the most important consequences of SC growth factor synthesis is to trigger their

neuregulin release from axons, discussed next.

Neuregulins

Neuregulins (NRGs) are intimate signals from axons to SCs. They are

important during development, during SC maintenance, and at several stages of

regeneration. NRGs arise from alternative splicing of the NRG1 gene and belong

to a family of growth and differentiation factors. These splice products have

included previously identified proteins called neu differentiation factor (NDF),

heregulin, acetylcholine receptor inducing activity (ARIA), and glial growth

factor (GGF). All have an EGF-like domain that activates erbB tyrosine kinase

receptors, and several have extracellular heparin-binding, immunoglobulin-like

domains. ProNRG, expressed as a transmembrane protein is cleaved, released,

and binds to heparan-sulphate proteoglycans in the extracellular matrix.

NRG is a “bridge” that links altered behavior in SCs to subtle axon injury or

frank axonal retraction. Thus NRG is released by axons from its proNRG axon

transmembrane precursor as a result of exposure to neurotrophins released by

SCs. PKC activation within the axon is involved in the release process [172].

Moreover, the sensitivity of axons to growth factor-NRG release may depend on

the type of axon and the type of growth factor. For example, NGF and GDNF

trigger NRG release from sensory neurons, whereas GDNF and BDNF trigger

its release from motor neurons. NRG accomplishes its actions using receptors

erbB2, erbB3, and erbB4 [101,211,464,480,582,745]. In SCs, erbB3, or erbB4 NRG

receptors are usually complexed to erbB2 as heterodimers. In fact, erbB2 does not

directly bind to NRG, but it signals NRG ligation through its association with

the other receptor isoforms. These involve activation of both the MAPK and

PI3K growth and survival pathways to facilitate subsequent SC survival, prolifer-

ation, and later myelination [80,173,242]. NRGs also play roles in early Wallerian

or Wallerian-like degeneration (see Chapter 3) to break down myelin and they

later provide information about axon size to SCs during decisions about


remyelination [470]. Once an axon attains a given caliber, initiating myelination

by its SC is a function of NRG signaling, in conjunction with other local factors

such as BM–integrin signaling. Questions about the role of erbB2, however, do

persist. For example, erbB2 null mice do not have deficits in SC proliferation

after nerve injury and their myelinated axon structure is preserved [27]. It seems

likely, however, that erbB2 signaling does overlap with that of other molecules,

and that these alternative pathways may subsume similar tasks in its absence.

Other SC proteins influence NRG signaling. For example, caveolin-1 is a

structural protein of lipid caveolae raft domains that acts as a cholesterol

binding and shuttling protein, and inhibits erbB2 signaling. Caveolin-1, in turn

may be upregulated by NGF binding to p75 in SCs, a complex pathway capable of

modifying SC behavior [665].

Retrograde loss of neurons

Following transection or axotomy of a peripheral nerve, retrograde loss

of motor, sensory, and autonomic neurons occurs. While such loss is less pro-

nounced in adults than neonates, it nonetheless reduces the pool of parent

neurons and axons available for reinnervation. When sensory neurons degener-

ate and disappear in ganglia, they leave behind a distinctive collection of peri-

neuronal cells that appear to proliferate around and within the site formerly

occupied by the neuron. These tombstones of sensory neurons are known as

“nests of Nageotte.” For motor neurons, loss is associated with local microglial

activation.

Retrograde loss is thought to occur because an axotomy deprives the neuron

of retrogradely transported, target-derived growth factors. This idea is the main

tenet of the “neurotrophic hypothesis.” A large number of growth factors includ-

ing the NGF-related neurotrophin family members [365], but also many others

(see Chapter 9), are capable of rescuing retrogade loss.

There has been controversy over the presence and extent of retrograde loss

of peripheral neurons after axotomy. Identifying true retrograde loss of sensory

neurons requires rigorous methods for counting, such as the physical or optical

dissector [117]. Complicating this issue is that some retrograde loss may occur

early after axotomy [118], but there may be more extensive ongoing retrograde

loss over time. There is less information available about chronic neuronal

survival after axotomy [239]. Kuo et al. did not observe retrograde loss of L4 and

L5 sensory neurons ipsilateral to a sciatic transection in adult rats at 2 weeks

but there was 10%–15% loss by 4 weeks from apoptosis [239,365]. Moreover,

the apparent loss in numbers later “disappeared,” a phenomenon attributed

to neuronal replacement. Coggeshall et al. [118] noted no loss of dorsal root

119Retrograde loss of neurons

myelinated axons, reflecting the number of parent sensory neurons in the

ganglia, out to 8 months after axotomy from cut or crush. Loss of unmyelinated

axons was noted, however, suggesting an interesting selective process of retro-

grade degeneration. Higher proportions of retrograde neuron loss have been

noted with more proximal lesions, perhaps from interrupting a greater propor-

tion of axons entering a given ganglia [132,764]. In the case of motor neurons

[58], retrograde loss is generally uncommon in adults, but death can occur after

root avulsion and after axotomy of cranial motor neurons.

Thus, true retrograde loss of parent neurons after axonal damage is not

extensive in adult neurons. Their lesser retrograde loss indicates less reliance

on target tissue support. Instead, they rely on growth factors synthesized by

themselves or by neighbor neurons (autocrine) or by supporting cells in the

ganglia (paracrine). Another mechanism of adult neuron survival may be intrin-

sic resistance to apoptosis. Heat shock protein 27 (HSP27) is a mediator of resist-

ance to retrograde neuron loss and it is upregulated after adult axon injury [123].

After activation by phosphorylation, it operates as a molecular chaperone

that protects neurons by refolding altered proteins. HSP27 inhibits cytochrome

c-mediated apoptosis by associating with Akt, stabilizes mRNA, and stabilizes

the neuron cytoskeleton, actions that rescue neurons from diverse insults

[39,123,376,485,618]. It is also expressed in growth cones.

A number of other molecules help to protect neurons from retrograde loss.

Members of the p53 tumor suppressor family, in particular DNp73, protect adult

sensory neurons by acting downstream of JNK proapoptotic pathway activation

[719]. PI3K and PKC signaling mediate neurotrophin-independent survival of

adult sensory neurons [143]. Overall then, while frank retrograde loss of injured

adult peripheral neurons can occur, it is frequently circumvented. Surviving

neurons, however, undergo alterations in their phenotype and behavior in

response to injury, and these changes are considered next.

Regeneration and perikaryal phenotype

A cascade of cell body changes is associated with axon injury and regen-

eration, first molecular and then morphological [181,403,707]. These changes, in

turn, have a profound influence on the later behavior of regrowing axons and

are discussed next. In both CNS and PNS neurons, a large menu of changes in

transcription and protein synthesis represents their conversion from a stable

transmission status to a dynamic and plastic regenerative phenotype. For

example, sensory neurons upregulate genes that express tubulin, actin, and the

growth-associated protein B50 (GAP43). The genes that are altered in response

to injury during regeneration are known as regeneration associated genes (RAGs).


In many instances they synthesize proteins that may be expressed within or just

proximal to growth cones and that participate in axon growth [707]. RAGs are

substantially influenced by growth factors, a topic discussed in Chapter 9.

The cell body reaction

There are early morphological changes, known as the “cell body reaction,”

of the perikarya of neurons that have undergone axotomy. Lieberman’s classical

work describes these structural changes in detail [403]. Briefly summarized, the

most prominent include changes in cell volume with variable early swelling and

late (if no reinnervation occurs) atrophy, an early rise in nucleolar volume,

disintegration of Nissl bodies, and later displacement of the nucleus from its

usual central position to an eccentric one near the cell membrane. Other

changes include: increases in smooth endoplasmic reticulum, variable changes

in the Golgi apparatus, loss or contrary increases of neurofilaments, and mito-

chondrial hypertrophy. The changes in mitochondria are distinguished from

artifactual swelling or vacuolation that can occur as a result of harvesting and

fixation [400].

The classical retrograde axotomy reaction, known as “central chromatolysis,”

describes dissolution and peripheral displacement within the cell body of dark

basophilic Nissl substance. Nissl substance is a histological term that describes

rough endoplasmic reticulum containing RNA and that stains dark blue and

punctuate on hematoxylin and eosin stains. “Nissl bodies” are normally promin-

ent in the center of the neuron and near the axon hillock. Hence the term

“central” chromatolysis after axotomy refers to its peripheral redistribution.

Central chromatolysis is thought to peak between 1 and 3 weeks following

axotomy and usually recovers with reinnervation. In some instances it persists

for long periods of time following an axotomy injury that has not recovered.

When recovery does occur, there may be a hyperchromatic phase with closely

packed numerous Nissl bodies suggesting upregulaton of RNA content. In the

case of motor neurons, axotomy also results in loss of its dendrite arbor and its

synaptic connections [62–65,610].

Many of the morphological changes following axotomy may simply reflect the

impact of altered protein synthesis. For example, the downregulation of neuro-

filament that maintains the internal lattice of the perikarya may allow the

nucleus to move to a peripheral rather than central cellular location. Despite

the displacement of RNA containing rough endoplasmic reticulum, overall neu-

ronal RNA content and trafficking are thought to increase after axotomy [403].

There is increased amino acid uptake for new protein synthesis.

There are other interesting aspects of the axotomy reaction. For instance,

after axotomy of the central branch of sensory neurons, the cell body reaction

121Regeneration and perikaryal phenotype

is attenuated or absent [403]. This may indicate that the local microenvironment

that axons occupy plays a very important role in determining how the entire

neuron tree responds to an injury. Both central and peripheral branches are

essential to sensory transmission, but their differing and segregated behavior

indicates unique molecular features. Another important feature is that neurons

contralateral to an injury develop changes in neuron morphology and gene

expression. In the case of mRNAs, these may be quite striking. The existence of

contralateral change is interesting and may operate through spinal cord connec-

tions or through injury signals that circulate by the spinal fluid. From an experi-

mental point of view, contralateral effects are very important to control for

when examining the impact of nerve injury.

Signals, RAGs

Neuronal survival, perikaryal support of distal regeneration, and RAG

expression are closely linked. The perikarya has the capacity to recognize, often

very quickly, when their peripheral axonal branches are injured or impaired.

How an injury to a distal axon tells the central cell body, up to a meter away, to

change a vast repertoire of genetic programs is unclear. Some of the candidates

for inducing this change include a retrogradely transported protein, a kinase,

and a calpain cleavage product of an intermediate filament [405,550]. Initial

upregulation of intermediate early genes may be the trigger for the subsequent

change in protein expression of injured neurons. The intermediate early gene

c-Jun is a major component of the AP-1 transcription factor complex. Its acti-

vation by JNKs occurs in response to injury and was observed to rise in facial

motor neurons of adult mice after transection. Mice lacking c-Jun had defective

target tissue reinnervation, axonal sprouting, and attenuated expression of

other regeneration molecules (CD44, galanin, and a7b1 integrin) [569].

Following axotomy or sectioning of their distal branches, dorsal root ganglia

sensory neurons upregulate mRNAs of a number of molecules that include:

GAP43/B50 (growth-associated protein 43), CAP-23 (cytoskeleton-associated

protein-23), ta tubulin, ATF3 (activating transcription factor 3), PKCb (protein

kinase C b), MAPK, NPY (neuropeptide Y), SCG10 (superior cervical ganglia neural-

specific 10 protein), HSP 27 (heat shock protein 27), and neuronal NOS (nitric

oxide synthase). In contrast, they downregulate TrkA, the high affinity NGF

receptor, TrkC, the NT-3 receptor, p75, the low affinity NGF receptor, NfM, the

medium subunit of neurofilament, and CGRP (calcitonin gene-related peptide)

[28,123,179,443,581,685,707]. Galanin, normally expressed in 2%–3% of intact

ganglia sensory neurons, is observed in 40%–50% of neurons after distal

axotomy. Recent evidence suggests that this upregulation plays a significant role

during regeneration. For example, mice lacking galanin have an accelerated


developmental apoptosis of sensory neurons, whereas adult knockout mice

exhibit a 35% reduction in regenerative activity [282]. Specific neuron subtypes

within dorsal root ganglia, of course, may preferentially upregulate one RAG but

not another. The exact time course for changes of gene expression during

regeneration vary.

In facial motor neurons, there is a rise in the expression of actin and tubulin

proteins after axotomy and a decline in the NfM and NfL neurofilament subunits

[675]. Additional molecules downregulated by axotomized motor neurons, sum-

marized by Boyd and Gordon [58] include: choline acetyltransferase (ChAT),

acetylcholinesterase (AchE), vesicle-associated membrane protein-1 (VAMP-1),

NT-3, NT-4/5, and TrkC. Upregulation takes place for: BDNF, TrkB, p75, RET,

GFR-a1, IL-6, LIFR, CGRP, VAMP-2, GAP-43, and HSP27 [123]. Among all of these

changes in both motor and sensory neurons, a critical theme is the upregulation

of tubulin, a key component of the microtubule and growth cone, and down-

regulation of neurofilament subunits, structural proteins of stable axons.

Overall then, changes in axotomized neurons reflect the shift from a stable

maintenance phenotype to a regenerating one. Many more changes are likely to

be discovered. Some of the alterations may have significant signaling implica-

tions. The upregulation of the stress-activated transcription factor ATF3, for

example, may alter expression of other genes by binding together and acting

with other leucine zipper transcription factors as heterodimers. Neurons with

overexpression of ATF3 have enhanced neurite outgrowth [612]. The relevance of

other alterations, such as those involving intermediate filament proteins is less

clear. a Internexin, an intermediate protein that has scaffolding roles and is

discussed in Chapter 2, has only low level expression in intact motor neurons,

but dramatically rises following axotomy [454]. Peripherin, a related Class III

intermediate filament protein, has a rise in its expression in peripheral sensory

neurons after injury [519]. Another interesting molecule that appears to rise in the

nucleus of injured motor neurons is EXT2, a heparin sulphate-glycosaminoglycan

synthesis enzyme. Its rise correlates with increases in the heparin sulphate pro-

teoglycans glypican-1 and syndecan-1 [484]. EXT2 may have a role as a coreceptor

for FGF-2 growth factor signaling (see Chapter 9).

In the case of GAP43/B50, its rise in synthesis by DRG sensory neurons after

a distant axon injury is accompanied by a rise in the levels of HuD [14]. HuD is a

neuron specific RNA-binding protein that interacts with the U rich regulatory

element of the 30 untranslated region (30UTR) of the GAP-43/B50 mRNA and

delays its degradation. HuD protein and GAP43/B50 mRNA expression increased

together in DRG neurons peaking at 7 days after a nerve crush and lasted 3 weeks.

Moreover, both were colocalized to granules containing ribosomal RNA in the

cytoplasm of neurons. By interacting with specific mRNAs, HuD and related

123Regeneration and perikaryal phenotype

proteins may offer protection and a longer half-life of important regeneration

related molecules like GAP43/B50. They may also offer protected transport into

axons to effect local later translation of protein.

PI3K-Akt [70] is discussed earlier as a survival pathway that mediates trophic

factor support of neurons and blocks apoptosis. Within individual neurons,

however, it may be that PI3K-Akt activation also serves roles that promote growth

and differentiation (for review see [189]).

In convincing perikarya to change (e.g., upregulating PI3K or RAC1) in ways

that facilitate axon regrowth, perineuronal glial cells may play a particularly

intriguing role. Perineuronal satellite cells in DRG elaborate GDNF and other

growth factors for use by adjacent neurons [250].

Local axon signaling in regeneration

Axonal protein synthesis

At one time, all protein synthesis was considered to arise from perikarya

with subsequent transfer down axons by axoplasmic transport. Over the past

decade, however, several investigators have established that axons are capable of

their own protein synthesis locally [124,703]. Moreover, axon protein synthesis

may be influenced by signals from SCs [124].

That local protein synthesis may help direct events in growth cones and

regenerating axons is important news in regeneration. Local approaches to

bolster synthesis of proteins and facilitate growth cones advance could eventu-

ally be applied therapeutically. In axons separated from their cell bodies, protein

synthesis may sustain “normal” axon and growth cone behavior for several hours

before Wallerian degeneration ensues [710]. Eventually these isolated axons fail,

however, and undergo degeneration.

Proteins with critical localized roles in axon protein synthesis include p38 and

TOR (target of rapamycin), caspase-3 (apparently independent of its role in

apoptosis), ribosomal-P0, and eIF-4E (eukaryotic translation initiation factor-4E).

Similarly, proteasome-mediated protein degradative machinery can also be

identified: 20s proteasome core, ubiquitin, and ubiquitin–protein conjugates.

Local axon synthesis may be important during preconditioning. Preconditioning

refers to an enhanced regenerative response to injury in axons having under-

gone a prior injury. Both synthetic and degradative protein machinery, locally

synthesized, are increased in preconditioned axons to prepare them for their

augmented response.

Yet other examples of locally synthesized proteins by mRNA in axons

are: b actin, peripherin, vimentin, gtropomyosin 3, cofilin 1, heat shock pro-

teins (HSP 27, 60, 70, grp75, ab crystallin), endoplasmic reticulum proteins


(calreticulin, grp78/BiP, ERp29), ubiquitin C-terminal hyrolase L1, rat ortholog of

human DJ-1/Park7, g-synuclein, superoxide dismutase 1, peroxiredoxins 1 and 6,

PGK1, a enolase, aldolase C/ZebrinII [742]. NGF and BDNF increased b–actin,

peripherin, and vimentin mRNA transport into axons to facilitate local protein

production.

Signaling that relies on local protein production influences growth cone

behavior. Several examples have emerged recently. For example, the role of

exclusive local axon signaling has been elegantly demonstrated using fluores-

cent conjugates of growth factors. Cy3-NGF induces rapid changes in chick

embryonic sensory growth cones within 1 minute, while still confined to either

the membrane or the peripheral cytoplasm of the growth cone [669]. Similarly, as

a result of local signaling, the proteins synthesized at injury sites may be those

that shut down growth cones and restrain regeneration. Recently, Wu and

colleagues [749] have demonstrated that RHOA GTPase is synthesized, localized,

and functional in growth cones where it facilitates collapse in response to

inhibitory cues. Several other locally synthesized proteins are likely crucial for

growth cone behavior. For example, b-thymosin, a molecule that binds and

regulates G actin polymerization into F actin, is locally synthesized in Lymnae

(freshwater snail) axons where it constitutively inhibits neurite outgrowth [702].

How coordination of regenerative events occurs between local sites and the cell

body is uncertain. At this time, the evidence would suggest that both local axon

synthesis, local signaling, but also anterograde transport of new proteins

together orchestrate the response of growth cones. The relevant site of signaling

may also depend on the molecule being considered. Insulin, a potent neuron

growth factor with receptors on both axons and perikarya, is capable of promot-

ing distal regrowth, and phenotypic changes of axons when delivered centrally

(intrathecally) to access perikarya. Similar doses are ineffective when given

locally at the injury site [689].

While yet unproven, it is possible that growth cones are primarily responsible

for a rapid local response that depends on local cues. The central cell body is

brought into play depending on the extent and severity of the injury. Lesions

closer to the cell body generate more rapid RAG changes. Transection with

prolonged disconnection from a target involves a much more prolonged loss of

target contact than crush and may require much more extensive perikaryal

input. Perikarya thus receive retrograde signals from the injured axons that

serve as indicators of the type and nature of the lesion. Other retrograde mes-

sengers such as growth factors seem to serve a general alerting role. The peri-

karyon is then prompted to shift its priorities for synthesis. To direct local

machinery for protein synthesis, regulatory proteins are transported anterogra-

dely to the injury site. They influence the repertoire and activity of synthesis.

125Local axon signaling in regeneration

Central perikarya not only respond to events in the axons, but interact locally

with each other and with adjacent glia. Thus there is an overall instructive dialog

at several levels to help direct the fine tuning of an injury response.

An importin–vimentin–pErk retrograde signal

Recently, a complex retrograde kinase signaling mechanism involving

an injury-induced upregulation of importin proteins in axons has been

described. Importins interact with dynein, the retrograde transport motor of

axons, and act as soluble transport proteins. They “normally” mediate transloca-

tion through the nuclear pore complex of a cell but have roles in axons as

well where they have been identified. Regeneration-modulating proteins

may employ a nuclear localization signal (NLS) that, in turn, binds to importins

for retrograde transport [252]. Thus, specific critical proteins, when complexed

with importin and dynein are capable of modulating cell body and nuclear

function. A related cascade of events involves vimentin, an intermediate

filament protein that is upregulated in sciatic axons after injury. Vimentin,

in turn, links pErk (phosphorylated and activated Erk1 and Erk2), members

of the MAPK signaling family, to importin b in a signaling complex that is

retrogradely transported [549]. This intricate story continues. Local axonal injury

may promote association of pErk to vimentin through a calcium signal. When

bound to the vimentin scaffold, pErk is protected from deactivation by

local phosphatases and is transported retrogradely with importin to signal

within the nucleus.

Once at the cell body, pErk is released from vimentin and signals widely

by directly phosphorylating transcription factors. pErk may inhibit cAMP phos-

phodiesterase and allow sustained rises in cAMP levels in the perikarya. Rises in

cAMP, in turn, mimic the action of a conditioning lesion [503,564]. This overall

mechanism also could explain why lesions of axons closer to the perikarya

induce more rapid changes in gene expression after axotomy. Proximal lesions

may facilitate more rapid and substantial retrograde signaling to the cell body

and nucleus.

Given the central signaling role of vimentin, it is not suprising that adult

sensory neurons in mice lacking vimentin had impaired regeneration.

Regeneration was restored by delivering calpain-treated recombinant vimentin,

a procedure that disrupts its polymerization into intermediate filaments. In

this specific series of interactions, therfore, it is “normally” polymerized

vimentin that is inactive during the cascade, and the depolymerized protein is

necessary to convey a signal or act as a transport scaffold. Interestingly, it is

a high local calcium intra-axonal environment that inhibits polymerization

of vimentin, in turn permitting its participation in retrograde signaling.


Vimentin intermediate filament is expressed only at negligible levels in intact

axons. Overall, the findings indicate an unexpected nonstructural second

role for an intermediate filament protein vimentin that is independent of

its role in the cytoskeleton. They also illustrate a remarkable conservation

of proteins for several tasks in the peripheral nerve. Vimentin and other

intermediate filaments may therefore act as motile scaffolds in a variety of

circumstances [549].

Roles of inflammatory cells and mediators

The context of axon regrowth is critical to its success and is determined

by neighboring cells and their actions. By day 3–5 following a peripheral nerve

injury, there is local hematogenous macrophage invasion into the proximal and

distal stumps of an injured nerve [236,580]. Brisk and abundant influx of inflam-

matory cells appears critical to the success of peripheral nerve regeneration. As

discussed in Chapter 3, a major role for inflammatory cells is to accelerate

Wallerian degeneration and allow regeneration to follow.

There are also important direct interactions between regenerating neurons

and specific inflammatory molecules [417,580]. Examples include interleukin 6,

interleukin 1b, TNFa, RAGE, and leukemia inhibitory factor (LIF) [267,287,

489,781]. Interleukin 6 (IL-6) helps to prevent retrograde neuron loss. Its synthe-

sis is upregulated in injured sensory and motor neurons [488,489], in turn

triggered by a signal from mast cells. IL-6 is also capable of promoting axon

outgrowth over myelin inhibitors similar to that of a conditioning lesion [92].

BDNF may participate in IL-6 actions [487]. Interleukin 1b, acting through non-

neuronal cells in ganglia, enhances axon outgrowth [287]. RAGE also influences

neurite outgrowth. Thus, for example, sensory neurons grown on amphoterin

(a RAGE ligand) with a dominant negative RAGE had attenuated neurite

outgrowth [583]. Phosphorylation of STAT3 is involved in RAGE regenerative

signaling [583].

Both RAGE and TNFa are also triggers for NF-kB activation, a transcription

factor. Thus NF-kB rises in perikarya of peripheral sensory neurons and trans-

locates to their nucleus during activation after a peripheral axotomy injury.

Moreover, the TNFa induction of NF-kB appears important for cell survival

after injury [182]. ZAS3 is a zinc finger transcription factor protein that competes

for sites on target genes with NF-kB and thus can modulate its activity. In

contrast to NF-kB, ZAS3 declines after axonal injury, a response that allows

enhanced NF-kB action [750].

Galectin-1 is a member of the b-galactoside-binding lectins that is expressed

in motor and sensory neurons. When oxidized, it binds receptors on macrophages

127Roles of inflammatory cells and mediators

that in turn stimulate SC migration and axon outgrowth [285,286]. Oncomodulin

is a macrophage-derived calcium binding factor capable of promoting regener-

ation in retinal ganglion cells [766]. Many additional inflammatory molecules

relevant to peripheral nerve regeneration likely exist, but studies have been

limited or have emphasized CNS actions [184].

Later clearance of macrophages from the regenerative milieu is also import-

ant since some of their products, such as NO and other free radicals, if

overexpressed, might collapse regrowing axons. Fry and colleagues [198] demon-

strated that the myelin of regenerating axons signal the Nogo receptors NgR1

and NgR2 on macrophages to stimulate their efflux from injured nerves. Nogo

signaling is better understood during inhibition of axon outgrowth and acti-

vation of RHOA, but this work demonstrates another role for these molecules in

the complex regenerative milieu. While macrophages are essential to Wallerian

degeneration and subsequent axon growth, prolonging their presence after they

have accomplished these tasks may be detrimental.

Collateral and regenerative sprouting and pruning

Axon “sprouts” can be of different types. Regenerative sprouts form

when an axon is damaged by transection or crush. Sprouts can form either at

the injured end of the axon, or more proximally. In the case of myelinated axons,

sprouts tend to form at the next most proximal intact node of Ranvier. Sprouts

can also develop from intact axons when they are adjacent to target tissues that

have been denervated. Branches from intact axons grow out to reinnervate all,

or part, of the denervated territory, a process termed “collateral sprouting.”

Diamond and colleagues identified important differences in the requirements

of NGF for collateral, but not regenerative, regrowth of cutaneous sensory nerves

[139–141]. Despite these intriguing findings, less overall attention has been

directed toward additional mechanisms that favor collateral sprouting in the

peripheral nervous system. It is possible that branching behavior in CNS cortical

neurons employs molecular mechanisms similar to peripheral collateral

sprouting. In the CNS, for example, netrin-1 induced axon branching by generat-

ing calcium transients. These, in turn, activated calcium/calmodulin-dependent

protein kinase II (CaMKII) and mitogen-activated protein kinase (MAPK) to accel-

erate the growth of sprouts from otherwise intact neurons [668]. Arborization of

dendrites is promoted by PI3K-Akt-mTOR (“mammalian target of rapamycin”)

signaling [306].

Since cultured neurons undergo injury on harvesting, the mechanisms

of collateral sprouting are difficult to study in vitro. This is unfortunate since

collateral sprouting has profound importance in understanding recovery from


nerve lesions in humans. For example, after human diagnostic sural nerve

biopsies, a procedure that involves resection of a length of the nerve trunk, there

is a significant area of sensory loss on the lateral side of the ankle [678]. If

followed over time, however, the area of sensory loss gradually contracts from

its outside edges inward (centripetal recovery) (see Figure 4.6). Recovery most

likely occurs from collateral reinnervation by intact neighboring sensory axons

growing into the denervated skin. In other instances, collateral sprouting does

not seem to help clinical injuries, an observation that indicates unforeseen

barriers to recovery.

Collateral sprouting also reinnervates sweat glands in the skin. Navarro

and Kennedy [499] investigated its time course from sural or saphenous

axons in mice. Adjacent skin next to either nerve was denervated and the

function of individual sweat glands was assessed by measuring sweat droplet

imprints formed in response to an injection of pilocarpine. Collateral reinnerva-

tion of sweat glands, identified as sweating outside of the normal territory of

these nerves, began approximately 28 days after denervation. Young mice had

accelerated reinnervation, whereas in older mice reinnervation was slowed.

Collateral sprouting is important in the response to partial motor axon loss

discussed in the next chapter [568]. In disorders with chronic motor neuron

dropout, remaining motor units undergo considerable remodeling and enlarge-

ment from collateral sprouts. Plastic motor unit remodeling helps to maintain

more muscle function than might otherwise be possible.

Collateral sprouting explains why injury to a mixed (containing motor and

sensory axons) peripheral nerve can result in axons being sent to inappropriate

branches. For example, a motor neuron may send branches back down the

correct motor branch after an injury but may also project collateral sprouts to

an inappropriate sensory branch being sent to the skin (cutaneous). Brushart and

colleagues [71] have shown that motor axons preferentially travel to motor

branches despite the growth of collateral sprouts, a property called preferential

motor reinnervation (PMR; see Chapter 6). PMR increases over time as the

projecting motor axons reach their correct muscle targets and inappropriate

cutaneous axons are pruned away. The degree of persistent arborization may

depend on the accuracy of the pathfinding [578].

Removal of redundant regenerative sprouts, or pruning of improperly

projecting branches, helps to recapitulate the matching of parent axons to

targets that were present before injury. This activity involves active breakdown

of excessive branches rather than a simple process of withdrawal or retraction.

Remodeling by breakdown of redundant collateral branches has been observed,

for example, during development of the CNS in Drosophilamushroom bodies [728].

Pruning has been difficult to document during peripheral nerve regeneration

129Collateral and regenerative sprouting and pruning

because most models examine axons penetrating into distal stumps

undergoing Wallerian-like degeneration. Distal stumps, for example, after a

crush injury, contain an admixture of new regenerating axons and damaged

axons undergoing degeneration. In this milieu, it may not be possible to

differentiate new axons undergoing degenerative pruning from those degener-

ating as a result of the original injury. An exception involves axons regener-

ating across nerve transection gaps early after injury. New axons may be

accompanied by profiles undergoing apparent degeneration from pruning

before they have reached the distal stump [450].

Collateral reinnervation, pruning, and sprouting have implications for

chronic degenerative disorders of the peripheral nervous system and in severe

disorders where the inciting disease process (e.g., Guillain–Barre syndrome) has

ceased or been arrested. Functional regeneration is unlikely or even impossible

in disorders with severe proximal (or perikaryal) damage. Remarkable neuro-

logical improvement can occur, however, if residual axons capable of providing

axon sprouts are present. This has been demonstrated experimentally in long-

term studies of mice rendered diabetic with streptozotocin (STZ), a toxin that

destroys insulin-producing pancreatic b cells [335]. The diabetic mice develop

neuropathy with eventual drop-out of sensory neurons and loss of skin inner-

vation. If their insulin producing pancreatic b cells spontaneously recover from

STZ toxicity, their diabetes resolves but repopulation of lost sensory neurons

does not occur. The denervated skin, however, can be reinnervated by residual

sensory axons through collateral reinnervation. There are many chronic neuro-

pathies associated with irreversible neuron loss but some degree of recovery

might be available through collateral sprouting. Alternatively, inappropriate

pruning may limit this kind of recovery.

Nerve trunk architecture

One of the most striking changes in overall nerve trunk architecture

after injury is compartmentation, the formation of minifascicles, complete with

their own multilaminate perineurium [479]. These changes occur in the distal

stump during regeneration as axons elongate within them and mature. Mini-

fascicles are also found retrogradely in the proximal stump and in experimental

neuromas [755]. Why injured nerves form such minifascicles is intriguing. For

example, they never develop without axons within them. Morris et al. [479]

originally suggested that they are a response of axons in an injury zone strug-

gling to preserve some modicum of protection within their immediate micro-

environment. Like the intact endoneurium, minifascicles may offer protection

through a blood nerve barrier.


The full repertoire of molecular signals critical to fascicular organization in

peripheral nerves is undefined. One relevant pathway is that of Desert Hedgehog

(Dhh) protein, a member of the hedgehog family of secreted ligands identified

during morphological development. Dhh null mice have peripheral nerves that

develop abnormally and have persistent minifascicle formation [541]. Dhh sig-

nals through a complex that includes the transmembrane receptor components,

Patched (Ptc) and Smoothened (Smo). Ptc2 appears to be expressed in SCs. Both

it and Dhh are downregulated in distal nerve stumps following injury, subse-

quently rising during regeneration [29]. Despite the unusual nerve trunks in Dhh

null mice, the exact relationship between Dhh ligation, axons, and SCs during

injury and repair is unknown. Additional molecules such as members of the

semaphorin guidance protein family (Sema3F and Sema3A) and their receptors

[602] are thought to regulate fascicle formation during nerve repair.

Fibroblasts are important participants in nerve injury and repair. While the

role of “endoneurial fibrosis” is probably overstated (a terminology that does

not truly distinguish between SC proliferation, fibroblast population, or involve-

ment of other cell types), these cells nonetheless have an important role in

remodeling the nerve architecture. They are distinguished as stellate or fusiform

cells with little cytoplasm but prominent rough endoplasmic reticulum and

a variable Golgi apparatus. The nucleus is ovoid with only minimal peripheral

nuclear condensations. Fibroblasts are not surrounded by the basal laminae

that SCs possess, a feature that is evident by EM and can be used to distinguish

them. Some have cilia and some have coated pits. Like SCs, fibroblasts can

develop alterations of their rough endoplasmic reticulum with deposits of dense

inclusion material in cisternae. Overall, their numbers increase after an axonal

injury [479].

The perineurium undergoes plastic changes following nerve trunk injury. The

first is separation of its layers [479]. Degenerative changes may occur later

including a reduction in the number of layers, in tandem with compartmenta-

tion described above. Axons, SCs, and epineurial-like collagen fibrils can be

found between perineurial layers, a feature absent in normal intact nerve

trunks. Some perineurial cells participate in phagocytosis.

Mast cells are the tissue cousins of circulating basophilic leukocytes. Some

mast cells reside in intact peripheral nerve trunks specifically in the endoneur-

ium and perineurium. After injury, there is a rise in the numbers of mast cells

and increased mast cell degranulation within the peripheral nerve. While their

overall importance during regeneration is uncertain, their constituents include

histamine, serotonin, growth factors, and proteases. Degranulation of mast cells

during nerve injury likely contributes to rises in local blood flow and breakdown

of the blood nerve barrier [169,533,787,807,809].

131Nerve trunk architecture

Summary

A symphony of interactions characterizes the early regenerative

response of the peripheral nerve. The major elements involve organizing mem-

brane formation for new sprouts, reprogramming of the gene expression of the

entire neuron, local axonal synthesis, dynamic growth cone behavior, and

phenotypic changes of SCs that facilitate axons in outgrowth and guidance. All

of these elements must come together in a coordinated fashion for significant

regeneration to occur. New molecules and signaling ideas, understood at the

level of single growth cones, have become part of the orchestra. In particular, the

control of polymerization and depolymerization of actin and microtubules and

their relationship to Ras family GTPases have profound effects on growth cone

behavior. “Survival” pathways such as PI3K–Akt are also crucial to the story.

Collateral sprouting of axon branches from intact neurons differs from regenera-

tive sprouting and presents itself as a key alternative for self-repair of the

damaged peripheral nervous system.

Suggested reading

Armati, P. J. (2007). The Biology of Schwann Cells. Cambridge, UK: Cambridge University

Press [23].

Gordon-Weeks, P. (2000). Neuronal Growth Cones. Cambridge, UK: Cambridge University

Press [228].

Letourneau, P. C., Kater, S. B., & Macagno, E. R. (eds.) (1991). The Nerve Growth Cone.

New York: Raven Press [1].

Ramon y Cajal, S. (1928). Degeneration and regeneration of the nervous system.

In Cajal’s Degeneration and Regeneration of the Nervous System, ed. J. DeFelipe &

E. G. Jones. Oxford, UK: Oxford University Press [572].


6

Consolidation and maturationof regeneration

Regeneration is an ongoing process. Once axons form growth cones that

cross injury zones, navigate distal stumps and approach former targets, they

encounter additional challenges. The interaction of new axons with their targets

involves a new series of events with new molecular requirements. Similarly

axons are not fully effective the moment they reach their target. They must

grow in caliber and may need to myelinate. This chapter will address these

events.

Forming new nerve trunks after transection

After the transection of a nerve stump, the proximal and distal stump

retract apart because they are normally under some degree of tension. Retraction

leaves a gap. In some cases, the gap can be repaired by suturing nerve stumps

together, but this is not always possible. Early studies successfully connected

the stumps with conduits, tubes, or in one report, synovium propped open by

a stainless steel spiral. Various surgical innovations are reviewed in more detail

in other texts and sources [421,422,660,775]. They include nerve grafts, silicone

tubes, vein–muscle conduits, fibronectin mats, veins, synthetic longitudinal

filaments, collagen tubes, polymer biodegradable tubes, and others. With these

strategies, eventual complete reconstitution of a severed new nerve trunk from

the proximal to the distal stump can occur. This was initially described by

Lundborg and colleagues [424–427].

When proximal and distal stumps are connected by tubes, an eventual

connection therefore does develop as discussed in the previous chapter

[610,740]. In the first week, extracellular exudative fluid and a fibrin matrix clot

form within the tubes. The clot retracts from the inner walls of the tube and

133

forms a bridge across which axons and SCs navigate toward the distal stump.

In successful bridges, later consolidation of regrowth includes the formation of

small “minifascicles” that begin to consolidate in the center of bridge material.

With time (months), gaps can be reconstituted with axons and resemble their

original nerve trunks (Figure 6.1).

While axons seem to eventually find their way across transection gaps, SCs

have difficulty in identifying the correct route. This is illustrated by using

conduits with a choice of which direction to grow. Y-shaped conduits were

constructed and attached to the proximal stump offering axons or SCs two

possible destinations. When one of the Y ends contained the distal stump and

the other contained an unrelated tissue, axons preferred to grow in the direction

of the distal stump. In contrast, SCs exhibited no such preference and migrated

into both ends. Axons crossing excessive gap lengths, however, require support

from additional growth factors or other interventions to arrive at the distal

stump [428,610]. In rats, this distance limit is approximately 10–12mm. All of

these findings illustrate the requirement that axons and SCs closely partner to

provide properly directed growth.

Figure 6.1 Images of reforming peripheral nerve trunks growing through a conduit

connecting a transected rat sciatic nerve at 21 days. There is a central core of

myelinated axons surrounded by layers of connective tissue. The insets show a

macroscopic view of the bridge (lower left) and a higher power view of a regenerative

unit with myelinated axons. The images are from semithin sections, epon embedded

and stained with toluidine blue.

134 Consolidation and maturation of regeneration

Reinnervation of muscle

Reinnervation of muscle involves reconnection of a motor neurons to

their endplates, reconnection of g motor neurons to spindles but also regrowth

of sensory axons into muscle. The latter comprise several types of axons as well,

including unmyelinated nociceptive axons and large myelinated axons that

reinnervate muscle spindles. When regenerating motor axons approach their

target muscles, they identify previous endplates in order to re-establish neuro-

muscular junctions. Terminal SCs, also known as perisynaptic SCs, first cluster at

denervated endplates to facilitate reconnection. Once this occurs, regenerating

motor axon terminals are then guided to denervated endplates initially by

growing along old “SC tubes” from terminal SCs. Despite this form of support,

reinnervation rarely fully recapitulates the preinjured state. Remodeled motor

units and fiber type grouping, discussed below, become permanent fixtures of

reinnervated muscles.

In collusion with terminal SCs, reinnervating motor axon terminals “escape”

after innervating their original endplate to branch further and innervate

adjacent neuromuscular junction endplates. When reinnervation begins after a

partial injury, first, terminal hypertrophic SCs send out processes over a distance

of several microns to link adjacent denervated junctions. By next fasciculating

with terminal SCs of adjacent junctions, they form a linked network of neuro-

muscular junctions. As this process unfolds, new reinnervating motor axons

follow these SC processes so that any given muscle fiber may now be innervated

by several sprouts, a condition known as polyneuronal innervation. During

later maturation, as axon reconnections are made, polyneuronal innervation is

eventually pruned away.

Terminal SCs have other interesting features. They appear to “sense” synaptic

transmission and have rises in their internal Ca2þ levels during motor nerve

stimulation. They also have muscarinic and purinergic receptors and, if synaptic

transmission is interrupted, changes in their gene expression occur. Blockade

of motor terminals through botulinum toxin can induce sprouting of terminal

SC processes followed by sprouting motor axons. This property has led to the

hypothesis that motor axon activity, rather than the presence of axons alone,

plays a critical role in whether axons (following terminal SC processes) sprout.

Since terminal SC processes engulf or “cap” axons at neuromuscular junctions,

they are also well placed to have other roles, such as buffering extracellular

potassium. For reviews of the properties of terminal SCs, see papers from the

Thompson and Gordon laboratories [319,632,664].

Sprouts into partly denervated muscle therefore can arise from any of the

intact residual motor axons innervating a given muscle. Tam and Gordon [664]

135Reinnervation of muscle

described several forms of motor axon sprouts in the vicinity of neuromuscular

junctions. Sprouts originating within the original junction area are called

“ultraterminal” while those arising from the motor axon just before the junction

are called “preterminal.” Others arise from more proximal nodes of Ranvier. In

some instances, combinations of these patterns and quite complex arborizations

may be identified. The unique behavior of terminal SCs can therefore explain

several features of reinnervation in the context of partial denervation. First, it

illustrates the influence of SC processes in guiding fine motor axon sprout

behavior. As discussed in Chapter 5, axons seem to accomplish very little if they

are “naked”; rather, they are most influential when they are tightly controlled

and shielded by adjacent SC processes. If denervation is prolonged, however, and

an axon has not yet arrived at an endplate, the terminal SCs may eventually

abandon them (see Chapter 8). Without terminal SCs, axons that arrive later fail

to recognize or reinnervate endplates [319].

The behavior of perisynaptic SCs and motor terminals also explains how

reinnervating motor units enlarge and render distinctive histological changes

within target muscles. In normal muscles, fibers innervated by axons originating

from a single motor unit are dispersed through the muscle and are admixed with

fibers innervated by other motor units. After reinnervation, this pattern changes.

Fiber type grouping refers to the tendency for muscle fibers of a given motor unit

to cluster together, indicating the reinnervation of denervated neighbors

by sprouts. Clusters, or “groups” from a single enlarged motor unit develop

because of the close collaboration between terminal SCs and axons that migrate

to nearby denervated neighbors.

Since axons dictate the contractile properties of individual muscle fibers (i.e.,

slow or fast twitch), all fibers in one motor unit develop identical physiological

properties. These properties can also be predicted by histological techniques.

Certain types of staining identify specific muscle fiber types: slow twitch fibers,

also called Type 1 fibers, contain higher levels of aerobic–oxidative proteins, while

Type 2 fibers emphasize anaerobic–glycolytic metabolism and are known as fast

twitch. Further subdivisions, not examined here include Types 2A, 2B, and 2C. The

ATPase stain can be used to discriminate these fiber types. At high pH values (9.4),

the stain is dark (strong) in Type 2 fibers but not in Type 1. At low pH levels (4.3),

Type 1 fibers are dark and Type 2 fibers light. With a given histochemical fiber type

stain then, fiber type grouping can be easily recognized. Muscle fibers belonging

to one reinnervating motor axon unit are clustered together and have identical

histological characteristics. In contrast, normal muscles have a scattered, or

“checkerboard” pattern of fiber types indicating an admixture of motor units.

Glycogen depletion is another classical method used to demonstrate fiber type

grouping during reinnervation. This is based on the idea that high frequency


stimulation of a single motor unit metabolically activates the muscle fibers it

supplies, forcing the muscle fibers to consume their glycogen supplies to sustain

excitation–contraction. If muscles are rapidly harvested and then frozen and

stained for glycogen after stimulation, the fibers innervated by the stimulated

motor unit are identified by their loss of glycogen staining; muscle fibers of a

given motor unit are normally scattered throughout the muscle. A reinnervated

“group” can be recognized because adjacent fibers from the same unit are

depleted of glycogen after stimulation.

In a slowly progressive disorder involving loss of motor axons and denerva-

tion, there may be ongoing loss of motor axons, while remaining axons sprout

in order to compensate for them. Motor neuron disease is an example. Thus,

remaining motor units enlarge, but eventually they too may succumb from the

disorder. When an enlarged motor unit that has been compensating for denerva-

tion in turn disappears, the consequences may be catastrophic. Weakness once

characterized as mild may rapidly become severe. If large remodeled motor units

drop out during a progressive motor disorder, the change can be detected

histologically as “grouped fiber atrophy.” Atrophy, a response of muscle fibers

to denervation, may occur in groups of adjacent fibers if they had been reinner-

vated together at an earlier stage of the disorder. Thus, grouped fiber atrophy

(Figure 6.2) is a signpost of progressive loss of motor axons. Grouped fiber

atrophy contrasts with an acute denervating process where atrophic denervated

Figure 6.2 A photomicrograph of a sample of ileopsoas muscle from a patient

with a severe progressive axonal polyneuropathy showing grouped fiber atrophy

(arrows). (Reproduced with permission from [785].) See color plate section.


fibers are dispersed or scattered throughout the muscle because sprouting and

grouping had not occurred. Atrophy of single muscle fibers from denervation

can be severe and can result in gross muscle wasting. In the past, neurodegen-

erative disorders with muscle atrophy were called “wasting diseases.”

Reinnervation of muscles can be mapped by one or more electrophysiological

approaches described in Chapter 4. These include measures of CMAPs, force

measurements, motor unit counts (motor unit number estimates (MUNE)), and

motor conduction velocities. In addition, the disappearance of spontaneous

electrical discharges (fibrillations) of denervated muscle fibers detected by

needle electromyography indicates reinnervation.

CMAP amplitude correlates with both the numbers of innervated muscle

fibers and the number of motor axons innervating the endplate. During reinner-

vation, there is a gradual rise in the amplitude of the CMAP. In serial studies,

both the times when the CMAP reappears (the initial potential will be small,

reflecting reconnection of the first motor unit), reflecting the first connections

of regenerating motor axons to the endplate, and its growth in amplitude, are of

interest. However, because of sprouting, recovery of CMAPs may not accurately

reflect the numbers of reinnervated axons. Measurements of overall muscle

force, rather than that of single motor units, however, have the same caveats

that apply to CMAPs. An apparent rapid recovery in force may occur because

of sprouting rather than the addition of new motor axons. MUNE estimates

may help to solve this difficulty by strictly identifying how many motor units

are present.

Motor axon conduction velocities slowly rise after contact and reconnection

of regenerating axons with the endplate. The rises reflect axon regrowth

in radial diameter, myelination, and maturation of nodal properties. During

further maturation, increases in myelin thickness and internodal length also

take place.

More mature neuromuscular junctions acquire more “stable” neuromuscular

transmission as both the presynaptic terminals and endplates reform. This is

described as a greater “safety factor.” Unstable transmission can be detected

by protocols that are routinely used to detect human disorders of the neuro-

musucular junction such as myasthenia gravis (Figure 6.3). Early partial motor

denervation or reinnervation, like these disorders, also exhibits physiological

properties indicative of immature and unstable neuromuscular junctions.

Repetitive stimulation is a technique that involves supramaximal stimulation,

between 2 and 50Hz, of motor axons innervating a given muscle. Trains of 6–10

CMAPs are recorded and the amplitudes between initial and subsequent CMAPs

are compared. Normally, this form of repetitive stimulation does not impair

neuromuscular transmission and CMAP amplitudes are maintained throughout


(a)

Fibrillation potentials

Enlarged motor unitpotentials

SFEMG Repetitive nervestimulation decrementNormal

synchrony

Jitter , abnormalsynchrony

(b)

(c)

(d)

(e)

Figure 6.3 Electrophysiological changes associated with denervation and

reinnervation in humans. Runs of fibrillation potentials, recorded by

electromyography (EMG), arising from individual denervated muscle fibers are

shown in (a) (scale is 50mv/d; 10ms/d). Enlarged voluntary recruited motor unit

potentials, recorded by EMG, arising because of collateral sprouting during partial

denervation, are shown in (b), (c) (scale, trigger line and arrow are at 1mv/d and time

scale is 10ms/d; *indicates a motor unit potential). Unstable neuromuscular

transmission is shown by SFEMG in (d) and repetitive nerve stimulation in (e). SFEMG

records the synchrony of two adjacent muscle fibers from the same motor unit. Jitter

refers to loss of synchrony or variation in the timing of the second potential, shown in

the lower trace. The upper trace is normal with near complete synchrony between the

discharges (scale is 100mv/d; 0.5ms/d). With repetitive nerve stimulation, CMAPs

decline in amplitude because of progressive failure of neuromuscular

junction transmission by motor units (scale 2mv/d; 5ms/d).


the stimulation train. In neuromuscular junctions with a reduced “safety

factor,” however, the CMAPs progressively decline, since a proportion of the

stimulated units begin to fail. This is known as an electrodecremental response,

and declines in the CMAP of over 10% are considered abnormal. Single fiber

electromyography (SFEMG) is another approach used to examine neuromuscular

junction instability. In this technique, popularized by Stalberg and colleagues

[638], a needle electrode with a small, active recording site on the side of the

needle shaft is introduced into a muscle. Voluntary potentials can be identified

as arising from individual muscle fibers by filtering out activity from most

neighboring fibers. By careful electrode placement and slight voluntary acti-

vation, potentials from only two adjacent muscle fibers from the same motor

unit can then be isolated and their firing synchrony examined. Unstable neuro-

muscular junctions (reduced “safety factor”) have increased “jitter,” referring to

an excessive variation in the synchrony of firing. Thus, the adjacent muscle fiber

potentials have variability in their ability to fire simultaneously. Blocking,

a phenomenon in which one potential fires but the adjacent potential fails,

also occurs. A variation of the SFEMG technique involves stimulating the nerve

rather than using voluntary recruitment, an approach applicable to animal

models. Overall then, electrophysiological properties can be used to indicate

that neuromuscular transmission is unstable, with a reduced “safety factor”

reflecting immaturity of sprouted innervating axons or unstable axons undergoing

degeneration.

EMG recordings can also be used to gauge reinnervation. During reinnerva-

tion, there is repair of abnormal spontaneous electrical activity. Fibrillations

disappear first. Next, functional reconnection occurs, and single or multiple new

motor units can be activated by slight, voluntary contractions. Early during

reinnervation, motor units may be quite small with multiple peaks (polyphasia).

Motor units then gradually enlarge. If axon reconnection to the muscle end-

plates is complete, motor units regain their normal sizes. If, however, there is

persistent partial loss of motor units, residual motor unit sizes can exceed

normal; this is due to sprouting. In conditions with long-standing partial but

permanent denervation, such as previous poliomyelitis, they may enlarge up to

5–10 times the normal size, and form “giant motor units” (Figure 6.3).

Reinnervation of the skin

We do not understand the guidance cues and trophic factors that allow

precise reinnervation of the skin, or what barriers are overcome in doing so. It is

assumed that, once axons reach the skin, their full ramification throughout the

dermis and their invasion of the epidermis becomes possible. Nonetheless, some


curious barriers to full reinnervation can be identified. Reinnervation may never

be complete after nerve transections and abnormal function may persist [152].

For example, patients note considerable fluctuation in their sensation early

during recovery from neuropathies. Paraesthesiae, or tingling asleep sensations,

may be prominent, a development that may arise from abnormal ectopic

impules generated by immature axons. There may be variations in sensation

depending on limb temperature and there may be misrepresentation of sensation.

Innocuous sensations of heat or pressure that instead evoke pain are termed

allodynia. In patients with neuropathy, cold sensations may be misinterpreted

as heat [767].

Polydefkis and colleagues examined the time course for skin reinnervation in

human subjects given local topical treatments with the neurotoxin capsaicin

(0.1% cream over an area of 35�50mm) [559]. Capsaicin activates TRPV1 chan-

nels and can eliminate unmyelinated axons in the skin within 48 hours.

The mean rate of epidermal nerve fiber reinnervation was approximately

0.18 fibers/mm per day. The onset of regeneration, indicated by the reappearance

of epidermal axons, was apparent by about 1 month after treatment. Sustained

regeneration, however, was slow, over 100 days and up to 350 days in some

subjects. It is uncertain to what distance or depth capsaicin neurotoxicity

extends. The success of reinnervation in this study may have thus depended on

how capsaicin dissipates after delivery and whether deep parent axons were

damaged. Overall, these and subsequent studies indicate that skin reinnervation

and remodeling can be prolonged.

Wiberg and colleagues [739] examined skin reinnervation in humans after

hand reimplantation. Eight patients were studied at intervals of 26–81 months

after surgery to coapt and repair the median and ulnar nerves, whereas the

radial nerves were not repaired. Multiple approaches were used to examine

sensory reinnervation of the hands. Four patients of the eight had a return of

two point discrimination, six to light touch, one to pain, five to cold, and six to

sweating. SNAP amplitude recovery was less than 40% of values recorded from

their intact contralateral hands, and sensory conduction velocities ranged from

50%–90% of the intact values. Mechanical thresholds to von Frey hairs were

variable with most ranging over 150% of the intact values, indicating that a

greater stimulus was required to evoke a response. Temperature thresholds to

heat, cold, and heat-pain were also variable, but in the range of 150%–200% of the

intact side, again indicating impaired sensory recovery (higher threshold than

normal). While recovery of PGP 9.5 labeled axons innervating the epidermis

occurred in two patients, the remaining patients had only approximately 50%

of the expected numbers of epidermal axons. Axons labeled with other markers,

such as CGRP and VIP, were also reduced in most patients to varying extents.

141Reinnervation of the skin

Interestingly, the radial nerve territories, where repair had not been carried

out, were nonetheless found to contain axons labeled for PGP 9.5, CGRP, or VIP.

All of these axons, however, were severely reduced in number. Their presence

suggested robust collateral reinnervation by the adjacent reconnected median

and ulnar nerves since spontaneous reconnection of the unrepaired radial nerve

was unlikely. Taken together, these results in humans illustrate the substantial

delays in reinnervation of the skin and the inability of collateral reinnervation

to adequately compensate for these losses.

Analagous to the terminal SCs associated with the neuromuscular junction,

terminal SCs help to guide regenerating sensory axons toward their original

targets [152]. These SCs have different characteristics from the SCs of peripheral

nerve trunks. They do not proliferate after denervation, and they also have

somewhat different biochemical properties when denervated. While the signifi-

cance of these differences is uncertain, these SCs upregulate acid phosphatase

while downregulating alkaline phosphatase, ATPases, and oxidoreductases.

Terminal SCs also express p75 NGF receptors and L1 basement membrane

protein. For sensory axons transducing touch/pressure, there are additional

guidance cues to aid reinnervation. Merkel cells, or touch domes that mediate

slowly, adapting small receptor field touch/pressure sensations may be one of

these cues. Thus, a proportion (40%) of Merkel cells survive denervation to guide

sensory axons back to their original sites. It is also thought, however, that newly

regrowing sensory axons can form new Merkel cells [152].

There may also be differences in the regeneration rate of different types

of axons. C and Ad axons may reinnervate more effectively than large caliber sen-

sory axons. Pacinian corpuscles can be redundantly reinnervated by multiple

sensory axons, and Meissner’s corpuscles reappear deeper in the dermal papillae

than they were originally placed [483,770]. Pacinian corpuscles, Meissner’s

corpuscles, Merkel touch domes, and the pilineural complex associated with

guard hairs may all display persistent structural abnormalities despite reinnerva-

tion. Why different types of axons should regenerate differently is unclear, but it

is likely there are different guidance cues and differing distributions of growth

factors available to them. There may be unique physical constraints to the types of

targets that different kinds of axons must reach and what basement membrane

they need to associate with. During development, presumptive proprioceptive

neurons in vitro extend more robustly on laminin and fibronectin and preferen-

tially express a3–5b1 integrin receptors, whereas cutaneous neurons prefer

laminin and express higher levels of the integrin a7b1 [240]. The regrowth of

sensory axons may be more prone to error than that of motor axons

[284,609,720]. There is also the possibility that sensory neurons have differing

intrinsic regenerative capabilities. For example, IB-4 non-TrkA GDNF sensitive


sensory neurons in vitro have less regenerative potential. Their tardy behavior

was not corrected even if they were grown on laminin or were forcibly trans-

fected with a7 integrin and GAP43/B50 [381].

Navarro and colleagues [501] examined reinnervation of the footpad skin

after sciatic nerve lesions in mice. After a crush lesion, PGP 9.5 labeled axons

reappeared in dermal nerve trunks by 14 days with rapid subsequent reinnerva-

tion of the epidermis, Meissner’s corpuscles, and sweat glands. Sweat gland

reinnervation paralleled a return of functional sweating. By 6 weeks, however,

the density of epidermal axon reinnervation remained incomplete. After a more

severe nerve transection injury, the timing of reinnervation was further delayed.

Reinnervation first appeared by 35 days after injury. By 6 weeks, the epidermis had

only rare short axons innervating it. Sweat gland reinnervation had not recovered

fully. The time courses for reinnervation by axons expressing the peptides CGRP,

SP, and VIP after crush or transection paralleled that of PGP 9.5 labeled axons.

Tseng and colleagues [695] described reinnervation of the skin in rats

that underwent chronic constriction lesions (CCI) of the sciatic nerve. These

injuries, discussed in Chapter 3, are models of neuropathic pain and involve

the placement of four loose ligatures around the nerve. There is subsequent slow

strangulation and degeneration of the nerve. Not all of the axons degenerate in

this model, and small unmyelinated axons are relatively preserved. If the sutures

were removed by week 4 after the lesion, there was a slow rise in reinnervation

of the skin by 6 weeks and 8 weeks but only to less than half of the pre-injury

value. Reversal of behavioral abnormalities such as thermal hyperalgesia and

mechanical allodynia developed more rapidly and completely. Only rats in the

decompressed group recovered. Since CCI is a partial nerve injury model, it does

appear that preserved cutaneous axons can serve an important role in guiding

regrowing axons back to the skin. As will be described below, however, morpho-

logical and behavioral recovery do not necessarily occur simultaneously. This

work also illustrates that pain generation is a complex process and does not

strictly correlate with the presence or absence of regeneration. Full and complete

regeneration, however, usually extinguishes neuropathic pain.

Misdirection of axons

Misdirected axons can fail to reconnect to their target. This outcome is

as detrimental to recovery as if no axon regrowth whatsoever had occurred.

Surgical approaches can help to prevent axon misdirection. For example, sutur-

ing predominantly motor nerve trunks to their own distal stumps, or sensory

nerve trunks to sensory distal stumps leads to better outcomes than suturing

mixed nerve trunks [660]. This finding illustrates that mixed trunks offer axons

143Misdirection of axons

more opportunities to go astray or be misdirected, limiting the success of overall

regeneration. Myelinated axons may also inappropriately grow through path-

ways previously occupied by unmyelinated axons [629]. Which factors misdirect

axons away from fully innervating distal stumps is uncertain. Specific cues, some

from SCs, can help direct sensory, motor, or autonomic axons to appropriate

targets. Guidance also depends on the type of axon being considered and

the roadmap to their destination. For example, in one study using retrograde

markers, a higher proportion of regrowing axons arrived successfully back into

their original tibial nerve trunk after a sciatic nerve injury than arrived later

into their original distal skin territory [562]. There may be fewer successful cues

for correctly directed regeneration into the distal skin territory compared with

more proximal nerve stumps. It is not uncommon to identify sensory deficits in

human skin with nerve injuries that have failed to recover even after decades.

“Digit splitting” in human hands, with localized loss of sensation to only one

half of a ring finger from an ulnar nerve lesion may persist for long periods. Loss

persists despite the presence of intact innervation in the nearby other half of

the digit, innervated by the median nerve. In this instance, there is a long-term

failure of invasion of axons that are meant to collaterally sprout from one half

of the finger to the other.

Brushart and colleagues first described how regenerating motor axons

preferred to enter nerve branches traveling to muscles, rather than to an equally

available branch directed to the skin (cutaneous branch) [71]. This interesting

property is referred to as “preferential motor reinnervation” (PMR) and occurs

through active pruning of inappropriately directed motor axons heading down

the cutaneous branch [72]. PMR and overall motor axon regeneration are facili-

tated by exogenous electrical stimulation [7]. Simulation, in turn, is thought

to recruit endogenous BDNF that acts through TrkB receptors. Electrical stimu-

lation also may support PMR by upregulating HNK-1 expression. HNK-1 is an

acidic glycan carbohydrate moiety that is exclusively expressed by the compact

myelin, basal lamina, and cell surface of SCs associated with motor axons in

motor fascicles [162]. Thus, HNK-1 promotes preferential outgrowth from motor

neurons, but not sensory neurons [441].

While adhesion molecules are discussed in more detail in Chapter 10, I will

briefly outline some of their roles in PMR. Franz and colleagues [190] described

high levels of neural cell adhesion molecule (NCAM) and its polysialic acid (PSA)

moiety in motor axons targeting motor pathways. Mice lacking NCAM failed to

exhibit PMR and had less withdrawal of misprojected axons, fewer collateral

sprouts, and smaller fields of arborisation. PMR was also abolished in mice

with enzymatic removal of the PSA moiety. Alternatively, Saito and colleagues

suggested that sensory fascicles did not express HNK-1 but expressed NCAM on


SCs associated with unmyelinated sensory axons [592]. Overall then, while NCAM

is expressed both on axons and SCs, there are discrepancies as to whether

sensory axons of motor branches preferentially express it. Most studies agree

however, that HNK-1 is a motor-associated basement membrane molecule.

F-Spondin extracellular matrix adhesive protein is also more selectively enriched

in motor branches where it facilitates axon outgrowth [82].

Are new neurons formed in ganglia?

There is no current evidence that endogenous stem cells provide

ongoing replacement of lost motor neurons in adults. In sensory ganglia, how-

ever, this possibility has been considered. Devor and colleagues [136] suggested

that sensory neurogenesis was ongoing in young and adult rats. These authors

identified surprising rises in neuron counts within ganglia over time. Their

conclusions were disputed by others who criticized the counting approach. The

concept has recently resurfaced but has not escaped controversy. Kuo et al. [365]

described a rise in nestin-labeled lumbar DRG sensory neurons ipsilateral to

sciatic nerve injury in adult rats starting at 2 weeks and lasting out to 6 weeks.

The numbers increased by adding NT-3. While nestin can indeed be used

to identify new neurons in vitro, this molecule is also present in endothelial

and other cells and may not accurately reflect neural stem cells in vivo [495].

Cheng et al. [107] examined the uptake of BrdU, a nuclear marker of the cell cycle

entry, in DRGs. Populations of proliferating DRG cells were identified that

developed after axonal injuries of the sciatic nerve. Most proliferating cells

represented perineuronal satellite cells. These cells also exhibited hypertrophy

and heightened expression of glial fibrillary acidic protein (GFAP). Some actively

proliferating cells might represent stem cells, but this possibility has not been

confirmed. Cheng et al. [107] also identified rare, mature intact neurons with

BrdU uptake, suggesting ongoing replication. The question of whether mature

neurons divide remains speculative. It has recently been shown that neurons

destined for apoptosis can inappropriately enter the cell cycle (and take up BrdU)

prior to death [264]. In turn, apoptosis is recognized to occur in sensory neurons

as a retrograde response to axon injury. Overall, it therefore remains controver-

sial as to whether new sensory neurons can arise in ganglia. There have been

no investigations as to whether autonomic neurons are capable of replacement.

Support of consolidation

Neurotrophic support is discussed in Chapter 9. In the case of regrowing

motor axons, trophic support does not facilitate all aspects of regeneration.

145Support of consolidation

In fact, when targets are recognized, growth factors may delay proper neuro-

muscular junction development. Peng et al. [548] observed that a mixture of

BDNF, GDNF, NT-3, and NT-4 and interventions to raise intracellular cAMP levels

promoted outgrowth of xenopus motor neurons. All of these stimuli, however,

downregulated the processes of agrin deposition and Ach receptor clustering

when the motor neurons were cocultured with muscle fibers. Coculturing

normally promotes the formation of neuromuscular junctions, and the changes

in agrin and Ach receptors are cardinal requirements for muscle innervation.

SC-conditioned media restored the requirements and allowed neurons to switch

to a synaptogenic state, modifying the actions of the neurotrophin mixture.

Thus, the findings confirm that the SCs not only facilitate growth of axons, but

also guide axons in their attempts to form stable neuromuscular junctions.

SCs that support regrowing axons elaborate different menus of growth

factors, depending on the type of nerve they are associated with. Thus, for

example, denervated ventral roots expressed mRNAs for pleiotrophin and GDNF

and supported motor axons in preference to cutaneous axons. In contrast,

cutaneous nerves upregulated NGF, BDNF, VEGF, HGF, and IGF-1 and preferred

to support cutaneous axons. Interestingly, these menus are modified over time

if cross-connections among regenerating nerve trunks are made [280].

Myelination

Remyelination of peripheral axons during regeneration is thought to

recapitulate events during development. The reader is referred to striking and

helpful depictions of its progress illustrated in other sources [23,661]. Overall,

myelination is a unique biological process in which SCs initiate the synthesis of

large amounts of lipid membrane material upon contact with axons. Moreover,

since axons must acquire a significant radial diameter before myelination

begins, it is not just axon contact that is relevant. During development, the

critical diameter of an axon that induces myelination begins at approximately

2 microns. Both myelination and radial axon growth can proceed concurrently at

this point. Myelinated axon segments have a 1:1 relationship with the SCs that

myelinate them. Thus, individual SCs line up along a given axon and each is

responsible for an internodal segment of myelin. Any given internode is only

associated with one SC. This differs from CNS oligodendrocytes where a given

cell supplies multiple axons.

The lipid material from the SC membrane is spirally wrapped around the axon,

then compacted. Spiraling of myelin occurs from the side of the SC membrane

facing the axon, and this process slowly “drags” the rest of the SC around with

it [79]. Compaction involves narrowing of spiral membranes and extrusion of most


of the SC cytoplasm. The collapsed cytoplasmic surfaces of the SC membrane

adjacent to one another form the major dense line, whereas the apposed extra-

cellular faces form what is known as the intraperiod line. Compaction does not

occur at some sites including the periaxonal collar, the outer SC loop, the

paranodal loops, and the Schmidt–Lanterman incisures. The portion of the

SC cytoplasm next to the axon, or periaxonal collar, is termed the inner mesaxon,

whereas that on the outer surface of the SC is called the outer mesaxon. All of the

events of myelination have specific triggers. For example, P0, the most common

protein found in peripheral myelin, helps to form an adhesive lattice during

compaction. Its extracellular domain forms interactions with other P0 molecules

on the adjacent membrane, helping to form the intraperiod line. Its intracellular

domain also interacts with like domains on the opposite membrane side to form

the major dense line of myelin. Thus, components of P0 operate as adhesives

on both sides of the membranes.

Myelination is not simply a product of an SC and axon interaction. There are

important signals from the basement membrane that are critical to initiate

myelination. One signal involves an interaction between a heparin sulphate

proteoglycan glypican-1 and the a4(V) motif of Type V collagen [112]. Glypican-1,

present on the SC surface, is required for both a4(V) binding, and for incorporation

into the normal SC basement membrane. The interaction of these molecules,

both apparently synthesized by SCs, is a prerequisite for subsequent myelin-

ation, as studied in cocultures of embryonic rat sensory neurons and newborn

rat SCs. There are several examples of abnormal developmental myelination

in transgenic mice that result from failed interaction between basement

membranes, adhesion molecules, and SCs [112]. Mice with mutations in the

laminin a2, a4, or laminin g1 subunits, rendering abnormal laminin 2, 8, or both,

respectively, all had defective myelination. Similarly, disruption of b1 integrin

or dystroglycan impairs proper myelination.

Several other molecules play critical roles in the development and mainten-

ance of myelin. One player is Krox20 (also known as Egr2), a zinc finger protein

transcription factor. Mice lacking Krox20 fail to develop peripheral myelin.

Later, conditional knockdown of Krox20 disrupted mature myelin sheaths.

NRG, discussed in Chapter 5, is essential for Krox20 expression [486]. Further

transcription factors with related actions include Oct6 (SCIP/TstI) and Sox10.

Oct6 is translocated to the nuclei (where it plays out its role as a transcription

factor) of SCs that proliferate during Wallerian degeneration and early regener-

ation [330]. Mirsky and Jessen [475] provide a list of additional transcription

factors that influence SC behavior.

Growth factors have important impacts on myelination both during develop-

ment and in adults during regeneration. NGF acting through TrkA promotes

147Myelination

myelination in sensory neuron-SC cocultures [96]. Similarly, BDNF facilitates mye-

lination both during development and in the adult. Mice overexpressing BDNF

had thicker myelin with a less proportionate increase in axon caliber [683]. High

doses of recombinant human GDNF given to rats altered axon–SC units and

promoted myelination of normally unmyelinated axons [279]. These interesting

findings occurred in the context of enhanced SC proliferation in response to

systemic intraperitoneal injections of GDNF (10 or 100mg/kg) given to adult rats.

Normally, in the rat, unmyelinated axons occur in a ratio of approximately

6 axons per SC Remak bundle. After GDNF, this ratio changed to 1.5–3 axons

per SC Remak bundle, and a number of the small unmyelinated axons showed

early myelination. Thus adult SCs do alter their behavior in response to growth

factors and GDNF may play a role in determining when an axon switches from

being unmyelinated to myelinated. TGF-b1 also appears important for the main-

tenance of mature myelin: mice lacking this growth factor had abnormal

“honeycomb” myelin with expanded SC cytoplasmic collars [130].

Long-term indices of regeneration

Relatively few studies have examined how indices of regeneration in

models recover during long-term follow-up. For example, after sciatic nerve

transection in rats, recovery of function may be substantially delayed even up

to 6 months. The long-term outcome after injury is influenced by the degree

of injury and the time to repair (if carried out). The types of assays examined

may also differ in their analysis of long-term recovery. While it is assumed that

more accelerated early regeneration translates into more rapid and long-term

recovery, this may not necessarily be the case. If there is an unsupportive nerve

or target microenvironment (see Chapter 8), or an opportunity for slower axons

to “catch up,” the long-term outcomemay not improve despite more rapid initial

regeneration.

McDonald and Kennedy [332,451] examined mid-sciatic nerve transections

and repair with a tube conduit in adult rats and transections in mice, respecti-

vely. They found that small CMAPs reappeared from the interosseous foot

muscles by approximately 6–8 weeks and grew slowly thereafter. By 10 weeks,

CMAP amplitudes were less than 25% of pre-injury values in both models.

By 20 weeks, values still remained significantly below the normal range. Motor

nerve conduction velocities only recovered to 50% of control values by 20 weeks.

The findings indicate that the behavior of initial reconnected motor axons may

not predict the course of full recovery. In contrast, crush injuries recovered more

rapidly [332]. CMAPs reappeared in mice by 3 weeks and by 8 weeks were 67%

of values prior to injury. Conduction velocities improved more rapidly.


Analysis of myelinated axon repopulation of injured nerves identifies

surprising structural plasticity. For example, after a mid-sciatic crush injury

in mice, Kennedy and colleagues [332] identified rises in the number and density

of mouse tibial myelinated axons from 2–8 weeks after injury to values higher

than intact nerves. Their axons, however, were smaller in caliber and their

myelin sheaths thinner. These findings confirmed that individual parent

axons gave rise to more than one daughter axon. By contrast, with transection

injuries myelinated axon numbers and density had only recovered by about

75% by 10 weeks after injury and regenerated axons were small in caliber

with thinner myelin sheaths. The tibial nerve is mixed but contains a large

proportion of motor axons. The sural nerve, almost exclusively sensory, had

similar patterns of change. McDonald et al. [451] examined myelinated axon

repopulation of rat sciatic nerves and their branches after 20 weeks distal to

3–5mm gaps following sciatic transection and temporary placement of silicon

regeneration conduits. In the sciatic nerve distal to the injury and in all its

three branches, the number of myelinated axons exceeded those of the intact

nerve as in the mouse model above. The axons, however, were smaller in caliber

indicating immaturity in radial growth characteristic of multiple daughter axon

branches.

The important caveats about electrophysiological and structural indices of

regeneration also apply to behavioral techniques. For example, measurements

of recovery of grip strength or walking may not coincide. Reinnervation of

proximal muscles below the knee may allow some function of long digit flexors

and extensors that impact grip strength. Thus, after mid-sciatic transection,

McDonald identified the first evidence of recovery of grip by 8 weeks, whereas

none was found with a walking track until 16 weeks. This correlates with other

reports. Walking track analysis of functional foot reinnervation in mice after

transection by Yao and colleagues indicated very limited recovery out to 90 days

after injury, despite immediate resuture. Crush injuries recovered within

approximately 20 days. Dysfunction from contracture of the foot muscles is a

complication of prolonged nerve injury [759].

It is difficult to conclude that sensory recovery differs substantially from

that of motor function using electrophysiological or behavioral criteria. Again,

the types of assays all have important caveats in their interpretation. By 20 weeks

after mid-sciatic transection in rats, McDonald found that SNAPs and sensory

conduction velocities from fibers reinnervating the digits remained substan-

tially below control values. The morphology of the sural sensory nerve trunk

did not necessarily predict electrophysiological properties of their more distal

digital axons. Neither mean myelinated axon diameter nor the number of

myelinated axons correlated with the amplitude of the digital SNAP. McDonald

149Long-term indices of regeneration

also found that withdrawal latencies to thermal stimuli of the hindpaw

shortened somewhat by 4 weeks after injury. No further improvement was

observed. Mechanical hyperalgesia of the hindpaw appeared within only 1 week

of sciatic transection, a finding that most certainly reflects collateral sensory

sprouting from adjacent and intact saphenous axons. Hyperalgesia persisted for

20 weeks thereafter. Similarly, skin pinch tests applied to the lateral aspect of the

paw with a forceps (graded for flinching with or without vocalization) returned

rapidly after transection, also indicating surprisingly rapid collateral reinnerva-

tion. Overall, there did not appear to be a correlation between how many myeli-

nated axons repopulated the sural nerve with measurements of thermal or

mechanical sensation. Each kind of measure requires separate interpretation.

Recovery from forelimb lesions have had less attention. Galtrey and Fawcett

[208] tested regeneration and functional recovery in rats following combined

median and ulnar lesions above the elbow. Nerves underwent crush, transection,

and primary repair (immediate direct suture of the proximal to distal stump),

transection, and crossed repair (median to ulnar and ulnar to median), or

transection without repair. At 15 weeks following injury, crush lesions exhibited

near complete recovery of most indices of regeneration: the staircase test, paw-

prints, grip strength, horizontal ladder, von Frey testing, and cold sensitivity.

For most of these tests recovery of function was crush>primary repair>crossed

repair>no repair. Even with primary repair after transection, some tests, such as

cold sensitivity remained very abnormal by the 15 week endpoint (<15% of

normal). Behavioral tests that depend on more complex function and proximal

limb muscles recovered more fully. Central remodeling and adaptation to distal

deficits compensated for failed distal regeneration.

In conclusion, long-term studies of reinnervation demonstrate surprising

complexity. It is important to emphasize that apparent structural recovery may

not translate into behavioral improvement. Interpretation of recovery requires

consideration of the type of axon being tested and the assay for recovery that

is chosen. Comprehensive sets of testing are undoubtedly required for predictions

of the overall outcome. In motor fibers, early reconnection of proximal muscles

that operate the paw may allow earlier recovery of walking and grip. Similarly,

with sensory testing, there may be surprisingly rapid recovery because of collat-

eral reinnervation. A drawback of collateral reinnervation, however, is that it

may be associated with painful allodynia.

Plasticity of the central nervous system

The central nervous system responds to injury and regeneration of the

peripheral nervous system. This may occur in several ways. Motor neurons,


for example, have extensive dendritic arbors that extend throughout the anterior

gray matter horn of the spinal cord, allowing them to form synaptic connections

with interneurons or afferent neurons. After axotomy, there is remodeling of

these arbors and they undergo shrinkage and retraction. Similarly, there is a loss

of axon boutons that normally connect with these dendrites. Regrowth of the

arbor and reconnection to synaptic boutons occur with peripheral reinnervation

[62–65]. Even with long-term reinnervation, however, there are redistributions

of the type and localization of the boutons associated with motor neuron

dendritic branches [65,770].

The central nervous system offers representative topographical maps of

somatic sensation and motor function at several levels. Topographic representa-

tion (somatotopic) of sensation is found, for example, in the dorsal horn of the

spinal cord, thalamus, and cortex. By using electrophysiological approaches (e.g.,

magnetoencephalograpy) and high resolution imaging (fMRI), these maps can be

identified and examined in response to peripheral nerve lesions. Such studies

indicate that maps are highly plastic and change, sometimes rapidly, to alter-

ations in sensory input. Anesthesia of body parts shifts the map toward greater

emphasis on the nonanaesthesitized body. Similarly, a peripheral nerve lesion

may permanently alter the size and distribution of its sensory map in the cortex.

After the peripheral nerve lesions that accompany transection, phantom sensa-

tions of the missing body part may be associated with altered cortical or subcort-

ical sensory maps, sometimes generating pain. The sequence of cortical events

following a focal lesion of a peripheral nerve in a human, such as the median

nerve has been reviewed [421]. In the distribution of the cortex previously

representing the nerve territory, there is initial loss of activity, followed

by expansion of the representation of adjacent intact territories into it. These

newly occupied territories are refined over time. Restitution of the previous

cortical map can occur after a crush injury, if reinnervation is complete.

In contrast, transection injuries with permanent misdirection of axons are not

associated with eventual return of normal cortical maps. Instead, a mosaic of

discontinuous cortical representation zones develops progressively over time

[714]. Transplanted hands can re-acquire portions of previously shrunken cortical

maps [218]. There is less information about cortical reorganization of motor

maps from selected peripheral motor lesions. Since most nerves are mixed,

containing motor and sensory axons, it may be difficult to distinguish their

relative impact on the cortex.

The overall significance of cortical maps is that they may offer an opportunity

to offset “permanent” peripheral nerve deficits by central compensation. The

neurobiology that underlies cortical remodeling is not fully understood. One

might appreciate, however, its capability to offer insights into compensation

151Plasticity of the central nervous system

for peripheral lesions; patients with irreversible peripheral lesions may benefit

from improved function and less pain.

Summary

The consolidation and maturation of regeneration are at least as

important as early regenerative events in dictating recovery. At this time, major

limitations of understanding persist in how motor axons and sensory axons

reinnervate their targets over the long-term. Both collateral reinnervation and

central nervous system cortical remodeling may modify long-term benefits,

but neither completely compensate for severe injuries such as transection.

SCs remain important partners in later aspects of regeneration and specific

trophic molecules may help. Repopulation of parent neurons with stem cells

at the level of the anterior horn or ganglia may appear attractive but face

challenges identical to those of native neurons during reinnervation of distal

targets. Local events that influence how distal axons behave will play major roles

in functional reconnection, irrespective of whether they arise from old or new

neurons.

Suggested reading

Armati, P. J. (2007). The Biology of Schwann Cells, Cambridge, UK: Cambridge University

Press [23].

Lundborg, G. (2004). Nerve Injury and Repair. Regeneration, Reconstruction and Cortical

Remodeling. Elsevier [421].

Mirsky, R. & Jessen, K. R. (1999). The neurobiology of Schwann cells. Brain Pathology,

9, 293–311 [475].


perin

euriu

m

epin

euriu

m

endo

neur

ium

Figure

2.2

Anillustrationoftheoverallorganizationoftheperipheralnerve

trunkanditscompartmen

ts.(IllustrationbyScott

Rogers.)

Figure 2.11 An illustration of a primary sensory neuron that resides in a dorsal

root ganglia. The neuron is described as “pseudounipolar” with a single branch from

the perikaryon (cell body) that then divides into central and peripheral branches.

The perikaryon is surrounded by closely apposed perineuronal satellite cells.

(Illustration by Scott Rogers.)

Figure 3.1 An image of a teased myelinated fiber illustrating segmental

demyelination between two nodes of Ranvier (arrows). Demyelination accounts

for neurapraxic nerve injuries.

Figure

2.13

Anim

ageofDRGsensory

neu

ronslabeled

byim

munohistoch

emistrywithanantibodydirectedagainst

theheavy

subunitof

neu

rofilamen

t(red

).Thearrowheadpoints

toasm

aller

neu

ronwithless

prominen

tneu

rofilamen

tstaining(sometim

eserroneo

uslycalled

“neu

rofilamen

tneg

ative”).Theright(green

)panel

isthesamesectionwithtransm

ittedlight.Theinsetsh

owsasingle

largeneu

ronwitha

“unipolar”

process

(arrow)emergingfrom

theperikaryon.N

ote

alsothattheneu

rofilamen

tlabel

showsapatchyandcomplexnetwork

within

the

perikaryonthatis

thoughtto

determineoverallneu

ronalsh

ape.

(Bar¼

20microns)(Imagetaken

byChuChen

g,Zoch

odnelaboratory.)

Figure 2.15 An image of the skin of a normal mouse footpad labeled by

immunohistochemistry with an antibody directed against the axon marker

PGP 9.5 (green). Note the dense ramification of tortuous axons up from the dermal

plexus into the epidermis. The epidermis is also covered by thick callus.

(Image taken by James Kennedy, Zochodne laboratory and reproduced with


Figure 2.16 Images of neuromuscular junctions from the mouse hindlimb labeled

by immunohistochemistry with an antibody directed against neurofilament (NF200,

green) identifying terminal motor branches. The junction itself is labeled with

a bungarotoxin (a-bTx, red) that binds the presynaptic terminals. (Image taken by

Noor Ramji, Zochodne lab and reproduced with permission from [571].)

Figure 3.2 An illustration of early events following a peripheral nerve trunk crush

(axonotmesis). Distal to the center of the crush zone, Wallerian-like degeneration

begins with breakdown of myelin and axons. (Illustration by Scott Rogers.)

Figure 3.3 An illustration of early events following a peripheral nerve trunk

transection (neurotmesis). In the distal stump, Wallerian degeneration begins with

breakdown of myelin and axons. (Illustration by Scott Rogers.)

Figure 4.4 Images of DRGs retrogradely labeled with fluorogold (FG, blue) after a sural

nerve injury. The sections are colabelled with an antibody against neurofilament (red).

(Image taken by X.-Q. Li, Zochodne laboratory.)

Figure 5.1 Illustration of the sequence of events during early regeneration from

a nerve trunk transection injury. The figures from top to bottom show transection,

followed by axonal endbulb formation, early regenerative sprouts, and bridging of

the transection zone. (Illustration by Scott Rogers.)

Figure 5.2 Images of axonal endbulbs using immunohistochemistry to label their

specific constituents. The panels on the left show discrete (arrow) and irregular

(arrowhead) endbulbs proximal to a chronic constriction injury lesion that

contain the mu opioid receptor, double labeled with an antibody directed against

neurofilament (Nf ). The panels on the right show CGRP containing endbulbs

proximal to a sciatic nerve transection also double labeled with neurofilament.

(Images reproduced with permission from [694] and from [788].)

Figure

5.4

Illustrationofearlyaxonsproutingfollowinganerve

trunkcrush

injury.Note

thateach

parentaxoncansendmultiple

daughter

sprouts.(IllustrationbyScott

Rogers.)

Figure 5.5 Images of axonal sprouts after a crush injury of the rat sural nerve

using immunohistochemistry to label CGRP. CGRP is displayed prominently in early

regrowing axons. (Image taken by X.-Q. Li, Zochodne laboratory.)

StableMyelinating

SCs GFAPerbB2

activationC-J un

activation

p75

DedifferentiationProliferation

Figure 5.11 Illustration of changes in SC phenotype after nerve injury. (Illustration by

Scott Rogers.)

Filopodia

Lamellipodia

ActinMicrotubules

Figure 5.7 Illustration of axon sprouts and a growth cone with its main components

labeled. (Illustration by Scott Rogers.)

Figure 5.9 Simplified drawing illustrating how growth cones change the direction

of their trajectory in response to localized gradients of growth factors such as NGF

acting on their TrkA receptors (green). RHOA GTPase (red) is thought to mediate active

repulsion and may be localized to growth cone membranes that are collapsing and

not advancing whereas RAC1 facilitates outgrowth and may be localized to the

membranes of advancing parts of the growth cone.

Figure 5.10 Examples of complex in vivo growth cones found at the proximal stump

of a transected rat sciatic nerve and labeled using immunohistochemistry for

neurofilament (red) or PGP 9.5 (green). Note that the appearance is substantially

different from growth cones studied in vitro. The PGP 9.5 labels the full extent

of the structure, whereas neurofilament is only found in portions of the growth

cone and in its proximal stem. (Bar ¼ 20 microns) (Reproduced with permission

from [450].)

Figure 5.12 Immunohistochemical images of a DRG proximal to a sciatic nerve injury

carried out 5 days earlier labeled with GFAP (glial fibrillary acidic protein). GFAP labels

glial cells including perineuronal satellite cells. These cells, closely apposed to and

surrounding sensory neurons, become hypertrophic and upregulate their content of

GFAP after an injury (arrows). (Images taken by Chu Cheng, Zochodne laboratory.)

Figure

5.13

Illustrationsh

owingthesequen

ceofch

anges

insensory

neu

ronsandSCsafter

atransectioninjury.Note

thedevelopmen

tof

hyp

ertrophic

perineu

ronalsatellitecells.Neu

ronshave

peripheraldisplacemen

toftheirNissl

substance

andtheirnuclei.Invasionof

hem

atogen

ousmacrophages

accompaniesproliferationandded

ifferentiationofSCspreparingthewayfornew

axonoutgrowth.(Illustration

byScott

Rogers.)

Figure 5.14 Adult rat peripheral sensory neurons in vitro showing extensive neurite

formation. The neurons underwent a preharvesting preconditioning injury and are

examined at day 3. The arrow points to a collection of glial cells in the culture that

have attracted neurites to grow towards it. The image illustrates the tropic properties

of supporting cells during axon outgrowth. (Culture prepared by Sophie Dong,

Zochodne laboratory.)

Figure 5.15 Immunohistochemical images of axon (labeled with an antibody directed

against neurofilament) and SC (labeled with an antibody to GFAP) outgrowth at day 7

from the proximal stumps of transected rat sciatic nerves. The images in the top panel

illustrate extensive outgrowth close to the proximal stump and the images in the

bottom panel are further distally. Note the close relationship between axons and SCs.

(Images taken by Y. Y. Chen, Zochodne laboratory.)

Figure

5.17

Immunohistoch

emicalim

ages

ofaxonsem

ergingfrom

theproxim

alstumpofatransected

ratsciaticnerve

(connectedbya

conduit)atday7labeled

withanantibodyto

neu

rofilamen

t.Note

therapid

falloffin

axonnumbersasthey

pen

etrate

theregen

erative

bridge.

Theinsetandarrow

showsanaxonin

amisdirected,or“w

rongway”

trajectory.(Rep

roducedwithpermissionfrom

[450].)

(a)

(b)

(c)

Figure 5.18 Immunohistochemical images of fine axons at the most distal end of

regenerative outgrowth by day 7 following transection of a rat sciatic nerve

(connected by a conduit). The axons are labeled by antibodies to bIII tubulin and

neurofilament. The small arrow shows a double labeled profile. The arrowhead and

large arrow show a tubulin, but not neurofilament labeled axon profile that is highly

irregular and partly misdirected. The images illustrate that very fine distal axon

outgrowth at the regenerative front may not label with a neurofilament antibody.

(Bar ¼ 50 microns) (Image taken by Chu Cheng, Zochodne laboratory and reproduced

with permission from [450].)

Figure

6.2

Aphotomicrographofasample

ofileo

psoasmuscle

from

apatien

twithasevere

progressiveaxonalpolyneu

ropathysh

owinggrouped

fiber

atrophy(arrows).(Rep

roducedwithpermissionfrom

[785].)

Figure

7.1

Illustrationofthemicrovascularsu

pply

oftheperipheralnerve

trunk.(IllustrationbyScott

RogersbasedonFigure

3.4

by[421].)

Figure 7.4 An example of the experimental setup used to study nerve blood flow

in the rat using either a laser doppler flowmetry probe or a hydrogen sensitive

polarographic microelectrode. In this setup, vasoactive agents were tested by infusing

them into the arterial supply of the nerve from the contralateral femoral artery.

(Reproduced with permission from [801].)

Chickembryonic

tissue

Fragment ofmouse sarcoma

Figure 9.1 Illustrations of the original findings of Levi-Monalcini and colleagues that

led to the discovery of NGF. Axons emerging from explanted sympathetic ganglia grow

toward mouse sarcoma cells that secrete a gradient of NGF. (Based on the original

figures by Levi-Montalcini [389] and reproduced with permission from [782].)

Figure 9.2 Illustration of members of the neurotrophin growth factor family and their

receptors. (Reproduced with permission from [782].)

Sympathetic neurons

TrkA NGF

TrkA NGF

TrkB

TrkB

TrkC

BDNF,NT-4/5

NT-3

Nociceptive fibers

Mixed afferents

Motor neurons

Large diameter afferents

Figure 9.3 Illustration of the peripheral neuron subtypes responsive to members of

the neurotrophin growth factor family. (Reproduced with permission from [782].)

NGF

Incorrecttarget

Correcttarget

Correcttarget

Absent orremovedtarget

Death ofincorrectlyprojectingneurons

Axoninitiation

Contact withtarget tissues

Figure 9.4 Illustration of the neurotrophic hypothesis. Neurons from axons that

arrive at a target source of growth factor survive whereas those that fail to arrive or

grow to incorrect targets undergo apoptotic cell death. (Reproduced with permission

from [782].)

RTK

PIP2

PIP3

PTEN

Akt

Activated

Pl-3 Kinase

GSk3

GSk3

PDK1 mTOR

Akt

Forkhead

Forkhead Survival

Survival

IKKIKB

NF-kB

NuclearImport

IKB

p50 p50

p50

p65 p65

p65

FastTranscription

Caspaseactivator

Apoptosis

IKK

Figure 9.5 Illustration of the PI3K-Akt neuron survival pathway activated by growth

factor receptors containing a receptor tyrosine kinase (RTK) intracellular signaling

domain. (Illustration by Scott Rogers based on diagrams posted by EMD/Calbiochem

(www.emdbiosciences.com).)

Autocrine

Target tissue

Paracrine

Schwann cell

Retrograde transport

Figure 9.6 Illustration of sources of growth factor support. (Reproduced with


Figure

10.1

Immunohistoch

emicalim

ageoftherelationsh

ipbetweenlaminin

(labeled

withananti-laminin

antibody,

green

)andaxons

(labeled

withananti-neu

rofilamen

tantibody,

red)in

aperipheralnerve

stump.Laminin

isassociatedwithbasemen

tmem

branes

of

SCssu

rroundingaxonsandwithbloodvessels.(Bar¼

50microns)(Imagetaken

byChuChen

g,Zoch

odnelaboratory.)

Figure 10.2 Immunohistochemical image of outgrowing axons (labeled with

an anti-bIII tubulin antibody, red) at the regenerative front of a transected rat

sciatic nerve (connected by a conduit) closely associated with deposits of laminin

(labeled with an anti-laminin antibody, green), likely laid down by leading SCs.

(Bar ¼ 50 microns) (Image taken by Chu Cheng, Zochodne laboratory and reproduced

with permission from [450].)

7

Regeneration and the vasa nervorum

Peripheral nerves are living dynamic tissues that thrive on a nutritive

blood supply. The vascular supply of the peripheral nerve, termed “vasa

nervorum” participates intimately in regenerative events and influences their

success. There are important morphological and physiological differences among

microvessels that supply nerve trunk, ganglia, and brain. Each entrains and

regulates its vascular supply from differing physiological perspectives,

depending on their need for metabolic support. Following injury, vasa nervorum

alter their behavior in unique ways that reflect their exposure to molecules

released within the microenvironment and that offer insights into the repair

process.

Blood flow and microvessels of intact nerve trunks

Nerve trunk blood vessels, or vasa nervorum, are supplied by upstream

arterial branches of major limb vessels. Sometimes these arteries and nerves

course together as neurovascular bundles. Peripheral nerves also share their

abundant blood supply with other structures in limbs such as bone, connective

tissue, skin, and muscle. For this reason, major ischemic lesions are likely to

target several tissues and cause widespread damage. The redundant and abundant

blood suppy of nerve trunks, however, can be advantageous because the interrup-

tion of a single artery is unlikely to cause significant ischemia. There are some sites

where there is ischemic vulnerability, known as watershed zones. These are found

at areas supplied by terminal branches of overlapping arterial trees. For example,

a nervewatershed zone has been identified in the proximal tibial nerve of rats [458].

While less readily demonstrable in humans, an analogous area in the proximal

sciatic nerves likely is present. The central, or “centrofascicular” portion of some

153

nerve trunks may also be a vulnerable type of watershed. If the ischemia is

prolonged (>3 hours), axons may be injured and undergo Wallerian-like degener-

ation. Centrofascicular damage is uncommon in large multifascicular nerve

trunks, suggesting that a more complex arrangement of their blood supply deter-

mines what parts may be at risk. In these nerves, ischemia insteadmore commonly

renders unpredicatable patterns of multifocal damage that depends on its severity

at individual sites within the nerve trunk [155,786].

Peripheral nerve trunks have an overlapping blood supply from microvessels

that form multiple connections, or anastamoses (Figure 7.1). The overall pattern,

as routinely observed by surgeons exposing nerves, is that of an extensive and

complex vascular plexus in the outermost epineurial layer of the peripheral

nerve trunk. Connections are found among arterioles or venules or between

arterioles and venules, where they are known as arteriovenous (AV) shunts.

AV shunts are found primarily in the epineurial plexus but they have also been

localized to the endoneurium. By direct examination it may be difficult to

distinguish AV shunts, arterioles, and venules in the epineurial plexus. Nonethe-

less, their existence has been demonstrated in several ways. By injecting micro-

spheres into the arterial vascular supply of the nerve, experiments have shown

that AV shunts allow microspheres to bypass capillaries to the venous system

and eventually deposit in the pulmonary vascular bed. Mathematical modeling

based on experimental measurements of nerve blood flow also predicts their

Figure 7.1 Illustration of the microvascular supply of the peripheral nerve trunk.

(Illustration by Scott Rogers based on Figure 3.4 by [421].) See color plate section.

154 Regeneration and the vasa nervorum

presence [368,369]. Why such shunts develop, or how they might contribute

to nerve trunk integrity is unknown. Redundancy in the vascular supply of the

nerve trunk, however, explains why long segments of nerves can be “mobilized”

by surgeons with relative impunity.

Vasa nervorum can be assigned into either an epineurial vascular plexus

or an intrinsic endoneurial plexus. The structure and physiological properties

of blood vessels from these two compartments differ, despite the fact that they

are anatomically connected. Epineurial blood flow, supplied by extrinsic arteries,

is ultimately responsible for downstream blood flow in the endoneurial compart-

ment. Feeder arterioles supplying the endoneurium arrive directly from adjacent

segmental penetrating arterioles. Arterioles within the endoneurium may also

be longitudinal in orientation, arising from penetrating vessels at remote sites

and traveling parallel to axons.

Different approaches to measure nerve blood flow highlight the differences

between epineurial and endoneurial networks. For example, the epineurial

plexus entrains higher levels of blood flow, exhibits prominent AV shunting,

has innervation of its arterioles, and has a leaky blood nerve barrier. In contrast,

the endoneurial vascular supply is constituted largely, though not exclusively, of

noninnervated capillaries that respond passively to blood flow changes. The

capillaries are associated with pericytes, smooth muscle-like contractile cells,

but their influence on local blood flow is uncertain. Endoneurial capillaries are

also somewhat larger in luminal caliber than those of other tissue beds [37].

Arterioles in the epineurial plexus are innervated by sympathetic adrenergic

unmyelinated axons that mediate local vasoconstriction. These vessels are simi-

larly innervated by axons containing peptides, called peptidergic (Substance P,

calcitonin gene-related peptide (CGRP)) that mediate local vasodilatation [20,137].

CGRP, in particular, is a highly potent vasodilator that is capable of relaxing

vascular smooth muscle through both nitric oxide dependent and independent

pathways [60,565]. Adrenergic and peptidergic innervation provide a form of

nerve trunk vascular self-regulation since they arise from their own parent nerve

trunk [574]. Moreover, both types of innervation are tonically active, and influ-

ence the luminal caliber of arterioles in real time. Since the epineurial plexus

feeds the downstream endoneurial compartment, this type of ambient sympa-

thetic and peptidergic control can thereby direct downstream endoneurial capil-

lary blood flow. For example, interrupting sympathetic or adrenergic activity

by pharmacological blockade or sympathectomy results in a rise in normal

endoneurial nerve blood flow. Alternatively, blockade of peptide receptors

(either CGRP or Substance P) is associated with declines in nerve blood flow

[795,798,801]. Discrete rather than diffuse zones of vascular regulation mediate

these changes because the overall innervation pattern of epineurial arterioles is

155Blood flow and microvessels of intact nerve trunks

likely nonuniform and segmental [801] (Figure 7.2). While uneven and segmental

adrenergic vascular responsiveness has been demonstrated, the peptidergic

innervation pattern has received less attention.

Hypercarbia is associated with a rise in cerebral blood flow in the CNS.

In peripheral nerves hypercarbia is associated with mild rises in epineurial

plexus blood flow but does not appear to influence endoneurial blood flow

[416,576,792]. The role of autoregulation also differs between the CNS and

peripheral nerve trunk. Autoregulation refers to the physiological maintenance

of blood flow during alterations in mean arterial pressure and it is a feature

of CNS blood flow. In peripheral nerve trunks instead there is an almost

direct linear relationship (possibly curvilinear) between endoneurial blood flow

and mean arterial pressure [416]. Explained differently, rises in mean arterial

pressure are accompanied by parallel passive changes in nerve blood flow and

constant levels of flow are not maintained. Both the absence of autoregulation

and overall lower levels of ambient blood flow in peripheral nerve trunks likely

reflect their lower metabolic requirements than brain. Normal endoneurial

blood flow ranges from 15–20ml/100g per min [416,588], values considerably

lower than the 80–120ml/100g per min in CNS gray matter. Blood flow in central

white matter is somewhat lower than that of gray matter, whereas ganglia

entrain higher blood flow than nerve trunks. Their vascular properties are

discussed next. Approaches to measure blood flow in small tissues like nerve

trunks and ganglia are also considered in detail below.

Figure 7.2 Image of an epineurial arteriole supplying the sciatic nerve in a

live anesthetized rat showing segmental vasoconstriction to topically applied

norepinephrine. The localized response (arrow) despite superfusion of the entire

preparation with norepinephrine suggests that there is variation in the density

of adrenergic receptors along these arterioles. (Reproduced with permission

from [801].)


Blood flow and microvessels of ganglia

Spinal dorsal root ganglia (DRG) are supplied from arteries that emerge

from segmental radicular arteries along the spinal column. These branches

anastamose with other branches arising from spinal arteries that supply the

spinal cord [3]. Unlike the nerve trunk, ganglia do not have an extracapsular

plexus comparable to that of the epineurium. Neurons in ganglia are promin-

ently placed in the subcapsular space. From this location, their axons coalesce

in the core of the ganglia before emerging into the dorsal nerve root. Few studies

involving ischemic damage to ganglia have been carried out and the vulnerability

of specific structures is unclear.

DRG have higher levels of blood flow than nerve trunks. Measuring approxi-

mately 30–40ml/100g per min [792], or 2–3 times that of the peripheral

nerve trunk, the higher flow values likely reflect a higher metabolic demand.

This difference parallels that of the CNS gray matter containing neuron cell

bodies and synapses that entrain higher levels of blood flow compared with

white matter. Signficantly, neither nerve trunks nor DRG are as well perfused

as spinal cord gray matter, where blood flow ranges approximately 50–60ml/

100 g per min and they are less well perfused than brain cortex discussed above

[808]. Oxygen tension values measured in ganglia are shifted to lower values

than those of nerve trunks, indicating heightened oxygen extraction. DRG

microvessels do not demonstrate significant vasodilation from hypercarbia

[796].

In sensory ganglia, unlike nerve trunks, partial autoregulation of ganglion

blood flow can be identified. Thus for mean arterial blood pressures between

80 and 120mmHg ganglion blood flow changes are less than expected and are

partly stabilized. One might also expect that, since ganglia have some autoregula-

tion, the influence of adrenergic vasoconstriction would be attenuated. Autoregu-

lation and adrenergic vasoconstriction can have opposite actions on blood vessels.

For example, when there are falls in mean arterial pressure from hemorrhage

or cardiac failure, rises in adrenergic tone compensate by increasing arteriolar

resistance, thereby increasing blood pressure but reducing flow to tissues. In the

DRG or CNS, neurons are highly vulnerable to ischemia and blood flow needs to

be maintained within a protective range through autoregulation and lessened

adrenergic input. Along these lines, DRG blood flow appears to have a negligible

input from adrenergic sympathetic fibers [796].

Blood flow within sympathetic ganglia is less studied, but it is likely their flow

would be comparable to that of DRG. One study suggested higher local flow than

sensory ganglia, but the measures may have been spuriously elevated by flow

through the adjacent carotid bifurcation [85].

157Blood flow and microvessels of ganglia

Measurements of nerve and ganglion blood flow

Accurate measurements of blood flow in small tissue compartments

like nerve trunks and ganglia are obviously challenging, particularly when experi-

mental models are used. Relatively few laboratories have used the techniques

listed here or have reported rigorously acquired results. Hopefully, newer imaging

techniques may soon provide accurate blood flow results within nerve comparable

with conventional results obtained over the last 20 years. Over this time, valid

classical approaches have included: quantitative microelectrode hydrogen clear-

ance (HC) polarography, laser Doppler flowmetry (LDF), [14C]idioantipyrene

distribution or autoradiography, and microsphere embolization. Detailed

reviews of the methods and their pitfalls have been published [415,783] and a

summary is provided below. Several related approaches that do not provide

quantitative blood flow data have yielded important corollary information. These

have included direct live videoangiographic imaging of the epineurial plexus and

its vessels [801] (Figure 7.2), quantitative morphometric measures of fixed or

unfixed peripheral nerves [654,806], andmeasures of indicator transit times [674].

Since different techniques may provide complementary information, combin-

ing them within specific studies has been a powerful way to confirm, supplement,

and extend critical findings. For example, the epineurial plexus is particularly

suited to sampling by LDF, whereas HC most suitably measures selective endo-

neurial blood flow measurements. Combining approaches is technically challen-

ging, but one can address interesting problems about how the two compartments

interact, for example, with pharmacological challenge or injury.

The technique of hydrogen clearance polarography (HC) is considered a “gold

standard” for quantitative measurements of nerve trunk and ganglion blood

flow. HC involves the detection of small currents from platinum microlectrodes

that have been polarized (hence polarography) to sense concentrations of H2. H2,

a highly diffusible gas, is delivered to tissues through an inspiratory gas mixture

in paralyzed, ventilated animals. HC microelectrodes must be small in caliber so

as not to disrupt the tissue they sample from, and they must also be linearly

sensitive to tissue concentrations of H2. Once the HC microelectrode detects

tissue saturation, or a plateau in the concentration of the hydrogen, the H2 is

shut off from the inspiratory gas mixture. The microelectrode then records

a washout, or clearance curve from which standardized approaches are used

to calculate endoneurial, or composite (weighing in the input of epineurial

blood flow and that of AV shunts) local blood flow. The calculations used in

this technique are provided in references [128,369,792]. It requires considerable

technical expertise, including rigorous physiological support and quantitative

characterization. Most laboratories with expertise in its use have used locally


manufactured linearly sensitive hydrogen microelectrodes with small diameter

tips of 3–5 microns. Some laboratories have reported use of smaller commercial

microelectrodes that have not undergone rigorous standardization or have used

very large microelectrodes that excessively disrupt the architecture of the nerve

trunk [288–290,520,642]. HC, when used with these caveats, offers quantitative

measures of selective endoneurial blood flow that can be performed serially

(Figure 7.3). Each measurement requires 20–30 minutes for the complete wash-

out of hydrogen, whereas rapid real-time variations in blood flow are difficult to

detect. Laser Doppler flowmetry, discussed next, allows real-time measurements.

Measurement of blood flow with laser doppler flowmetry (LDF) involves the

placement of afferent and efferent fiberoptic probes (usually combined into

one “probe”) over the surface of the nerve trunk or ganglion (Figure 7.4). LDF

utilizes the doppler principle, in which the efferent fiber emits a laser light

Figure 7.3 Examples of hydrogen clearance curves recorded from the endoneurial

vascular compartment of rat sciatic nerves using a polarographic microelectrode.

The clearance curves are monoexponential or biexponential. Note the more rapid

washout, indicating higher flow, in nerves exposed to capsaicin, an agent that

causes acute release of CGRP and SP, vasodilating peptides expressed in terminals

innervating epineurial arterioles. In contrast, the lower two curves exposed to

hCGRP (8–37), a CGRP antagonist, are flattened indicating slower washout and

reduced flow. (Reproduced with permission from [791].)

159Measurements of nerve and ganglion blood flow

signal to erythrocytes moving in its field of application. The afferent probe

measures reflected light, shifted in wavelength by the doppler effect, from

the movement of erythrocytes. Thus LDF is used to measure erythrocyte flux,

or the product of individual erythrocyte velocities and erythrocyte numbers. LDF

is sensitive to real-time changes in erythrocyte flux but vigorous care and

controls are required: (i) lighting conditions have a significant bearing on the

amplitude of the afferent signal and must be standardized, preferably with room

lights and other light sources turned off while the measurement is being made;

(ii) multiple sampling along the nerve trunk or ganglion is required, since very

slight shifts in the position of the probe can dramatically alter the recorded

signal. These kinds of variations are a result of the highly variable anatomy

of the epineurial nerve plexus and the lesser likelihood of being able to sample

uniform areas. In practice, measurements are best taken from several (a mini-

mum of ten sites) individual adjacent positions [315]; (iii) microtremor from the

hand of the experimentalist alters afferent signals, thus mandating the use of a

micromanipulator used to position and hold the probe; (iv) strict control of near

nerve temperature is required since local nerve temperature during exposure

Figure 7.4 An example of the experimental setup used to study nerve blood flow

in the rat using either a laser doppler flowmetry probe or a hydrogen sensitive

polarographic microelectrode. In this setup, vasoactive agents were tested by infusing

them into the arterial supply of the nerve from the contralateral femoral artery.

(Reproduced with permission from [801].) See color plate section.


and experimentation can vary greatly from central body temperature. Since LDF

preferentially samples from higher flow blood vessels adjacent to the probe, and

since deeper penetration into the nerve is limited, its measurements emphasize

flow in the epineurial plexus.

[14C]idioantipyrene distribution or autoradiography is also a gold standard

approach for the measurement of nerve and ganglion blood flow [459,588]. Iodoan-

tipyrine is freely diffusible among tissues and, after systemic administration,

its distribution can be measured in nerve. Its segregation into the endoneurial

or epineurial compartments can also be quantitated in specific regions with

autoradiography. Unlike HC, however, single and not serial measurements are

available in experimental preparations.

Microsphere embolization involves the administration of radiolabeled micro-

spheres through the bloodstream. Blood flow is calculated by using the distribu-

tion of their radiolabel to nerve samples. The approach has been criticized for

generating lower blood flow values in peripheral nerve trunks, and this may be a

technical limitation of the approach [561,662,680]. The reason for lower blood

flow measurements is that AV shunts found in vasa nervorum allow through

passage rather than capture of indicator microspheres, underestimating their

distribution. Other isotope distribution techniques have used [3H] desmethylimi-

pramine, and [14C] butanol [99,652] and do not suffer from this limitation.

Specific morphological techniques to examine vasa nervorum are available.

Standard sampling of nerves with fixation, embedding, and transverse sectioning

are used to appraise blood vessel structure, numbers, and density. High-quality

EM allows measurements of basement membrane thickening, smooth muscle

duplication, and other aspects of vessels [437]. While fixed tissues have also been

used to measure luminal areas of blood vessels, the approach obviously does not

reflect dynamic luminal caliber in physiological situations [603]. One technique

that may compensate in part for fixation problems involves perfusion of live

anesthetized preparations with a mixture of India ink, gelatin, and mannitol

[66]. After euthanasia and completing the infusion (identified by staining of

cutaneous vessels), the preparation is cooled at –5�C, allowing the intraluminal

mixture to congeal and harden. The nerves are then harvested, fast frozen in

molds of OCT (optimal cutting temperature compound), and cryostat transverse

sections made. Luminal profiles are outlined in black India ink. Such profiles,

obtained from unfixed tissues, can be counted and sized, both in the endoneur-

ium and the epineurium [805,806]. Finally, another approach uses fluorescein

imaging of blood vessels in live, anesthetized preparations or humans and can

provide information about vascular density and other properties.

Several methods are available for measuring oxygen tensions. One approach

uses oxygen-sensitive microelectrodes. Some of these microelectrodes operate

161Measurements of nerve and ganglion blood flow

through the principle of polarography, analogous to the approach discussed

above using hydrogen-sensitive microelectrodes. Since individual point meas-

ures of oxygen tension in tissue have questionable relevance, an appraisal of

tissue oxygenation requires the construction of oxygen tension histograms

sampled from multiple sites and depths [697,794].

Specific acute ischemic nerve injuries

Peripheral nerve trunks are much more difficult to render ischemic

than the CNS. Nerve trunks resist acute ischemia given their rich overlapping

vascular supply and lesser metabolic demands. Despite this resistance, however,

there are well-characterized clinical neuropathies that involve nerve ischemia.

Examples include compartment syndromes, discussed earlier in this text.

There are autoimmune conditions, known as vasculitis, involving inflammation

of blood vessels, including those of peripheral nerves. In these neuropathies,

damage can be highly variable, targeting one nerve and not another, in a pattern

known as “mononeuritis multiplex.”

There are some forms of focal diabetic nerve injury, resembling vasculitis,

that probably arise secondary to ischemia. For example, a type of diabetic

neuropathy known as lumbosacral plexopathy is thought to result from focal

ischemia of the lumboscacral plexus that supplies the lower limb. Leg paralysis is

the unfortunate result for patients [157,567]. Altered physiological properties

of the vasa nervorum in diabetes that develop even in early disease, contribute

to these types of injury. Defective vasorelaxation is a common theme and may

arise from impaired endothelial nitric oxide release, availability, or signaling.

Given these properties, diabetic nerves exposed to minor or innocuous levels of

ischemia are considerably more vulnerable to damage [512,513].

Several forms of experimental neuropathy are used to understand and model

the impact of ischemia on nerves. Multiple arterial ligation to the hindlimb results

in ischemic nerve damage, provided enough of the redundant vascular supply

is targeted by the procedure [129,355–359,458,605]. In these models, the mid-

portion of the sciatic nerve trunk is the most susceptible area to arterial inter-

ruption. The time window required for ischemia to cause permanent axonal

injury appears to be at least 3 hours, substantially longer than what is required

for CNS damage. After ischemia, reperfusion may account for additional ongoing

damage [605], a phenomenon referred to as reperfusion injury.

Ischemic neuropathy has also been studied after microsphere embolization of

the arterial supply of the nerve trunk, an approach described by Nukada and

colleagues [514–517]. In essence, this approach involved seeding the arterial

branches to the vasa nervorum with microspheres that occluded multiple

feeding vessels and caused ischemia. Within 12 hours following ischemia,


myelinated axons examined by LM became dark with or without light cores, then

developed swelling and secondary myelin thinning. Later myelin changes led to

overt demyelination and axons became “attenuated,” or atrophic. Overall, the ische-

mic changes were identified 12 to 48 hours after the ischemic insult but diminished

by 7 days. They were then followed by Wallerian-like axonal degeneration.

Other models of ischemic neuropathy have used intra-arterial infusions of

arachidonic acid or topical application of endothelin (ET) over a portion of the epineurial

vascular plexus [542,797]. Both of these agents cause severe vasoconstriction of

vasa nervorum. ET, the most potent vasoconstrictive agent known, generates

intense, albeit transient ischemia. A further form of nerve ischemia occurs in

the chronic constriction injury (CCI) model of neuropathic pain previously discussed

[41]. CCI is associated with ischemia from the placement of four loose ligatures

around a nerve trunk with gradual swelling and strangulation of the blood

supply [631].

Nerves rendered completely ischemic or hypoxic, however, do not demon-

strate immediate conduction failure or block! There is an important property of

axons known as resistance to ischemic conduction failure (RICF). RICF refers to the

property of nerve trunks under conditions of complete ischemia to nonetheless

retain their excitability for a period of time. In nerve trunks supramaximal

stimulation generates a NAP, the amplitude of which correlates with the number

of excitable axons that contribute to it. After complete interruption of their

blood supply in vivo, axons within a nerve trunk develop a gradual loss of

their capability to propagate the NAP. Thus, for a 50% decline in the amplitude

of a NAP, it takes 20–30 minutes of complete ischemia or hypoxia: this time

is referred to as the duration of RICF. Some conditions, such as diabetes, are

associated with a prolonged RICF. In contrast, in vitro excised nerves without a

blood supply, kept moist in physiological saline but exposed to air, retain their

excitability for much longer periods. Unlike nerves embedded deeply in a tissue

compartment, excised nerves are maintained by atmospheric oxygen.

After nerve trunks lose their excitability during complete ischemia, axonal

ischemic conduction block can be detected. Conduction block refers to a physio-

logical, rather than structural interruption of action potentials traveling along

the nerve trunk. Like the more commonly considered conduction block from

demyelination of internodes, ischemic axonal conduction block is identified

using electrophysiological techniques. Stimulation distal to the site of the block

identifies normal excitability with normal distally evoked NAPs or CMAPs.

In contrast, stimulation above or proximal to the level of ischemia is incapable

of exciting the distal nerve or muscle because the action potentials are blocked

across the involved zone. This phenomenon is occasionally captured in patients

with ischemic axonal lesions from vasculitis [314,543]. Short periods of complete

ischemia can induce transient conduction block without permanent damage.

163Specific acute ischemic nerve injuries

Longer periods of ischemia, however, render permanent damage to the axon.

After 3 hours of complete ischemia, Wallerian-like degeneration typically

develops and there is loss of excitability of the distal segment of the nerve with

disappearance of the CMAP evoked from distal stimulation (Figure 7.5).

ET (Left side)

Pre

2 hours

24 hours

48 hours

7 da ys

14 days

k

k

2 mv

1 ms

n kn

Saline (Right side)

Figure 7.5 An example of acute conduction block from ischemia followed by axonal

degeneration. The tracings are CMAPs recorded from the interosseous muscles of the

foot of a rat following supramaximal stimulation of the sciatic nerve at the sciatic

notch (n), or more distally at the knee (k). Between these stimulation sites, the nerve

was exposed to endothelin, a topical and potent vasoconstrictor on the left and carrier

alone on the right. Note that there is acute conduction block at 2 hours because the

distal nerve (k) remains excitable but stimulation more proximally (n) is blocked.

At 24 hours and thereafter, all excitability is lost because Wallerian-like degeneration

ensues from severe axonal damage at the ischemic site. This example is from a rat

with experimental diabetes. (Reproduced with permission from [789].)


To summarize then, axons within nerves in vivo take 20–30 minutes to lose

their excitability after complete ischemia because of RICF. After this period of

time, ischemic conduction block may be detected. If the ischemia is prolonged

beyond 3 hours, however (and the nerve is not excised and exposed to

air), permanent ischemic damage with later Wallerian-like degeneration will

ensue. In diabetes, RICF is prolonged out to 30–40 minutes, but beyond this

time, Wallerian-like degeneration develops more rapidly. The overall result

is that, despite prolonged RICF, diabetic nerves have greater susceptibility to

ischemic axonal injury [512,786,789].

It is likely that sensory ganglia in humans also undergo ischemic injuries but

they are not commonly characterized. Some acute forms of sensory “radicular”

neuropathy, as occur in diabetes, might arise from ganglion ischemia. It is also

possible that common degenerative spinal bony changes known as spondylosis,

with narrowing and compression of intervertebral foramina might cause nearby

ganglia ischemia. Ganglia are more vulnerable than nerve trunks to ischemia

because of their difference in metabolic demands.

Experimental models of sensory ganglion ischemia are available and like nerve,

diabetic ganglia are also more sensitive. For example, ET vasoconstriction after

topical application to ganglia caused ischemic necrosis of diabetic sensory

neurons, intraganglionic axon damage, and downstream degeneration of distal

axon branches [753]. Neurons disappeared and were replaced by nests of

Nageotte or had nuclear disruption identifying apoptosis. Alternatively, some

neurons had peripheral displacement of their nuclei and loss of neurofilaments

characteristic of a retrograde cell body response to damage, as occurs after axot-

omy. Thus, ganglion ischemia generated three separate pathological reactions:

ischemic necrosis of neurons, apoptosis of sensory neurons, and a retrograde

cell body response to axonal damage. The time window of ganglion vulnerability

to ischemia is unknown.

Local blood flow and the role of ischemia in acute nerve injury

Ischemia may not contribute to routine mechanical nerve trunk injuries.

Lesions like crush or entrapment could disrupt local microvessels through direct

shearing, compression from edema of the endoneurium, or venous engorgement

in the case of entrapment. Alternatively, the rich, overlapping anastamotic blood

supply of nerve trunks described above and the prolonged periods of intense

ischemia required to cause permanent axonal damage makes nerves resistant

to ischemia. There has been no clear demonstration that ischemia occurs or

plays a significant role in chronic entrapment or acute traumatic nerve injuries.

In these circumstances, it seems unlikely that a single compressive lesion would

165Local blood flow and the role of ischemia in acute nerve injury

disrupt the nerve vascular supply continuously for 3 hours or longer. Instead,

most types of chronic entrapment or compression likely damage nerves through

repeated mechanical injury. Unlike ischemia, local demyelination is prominent

in these lesions with later progression to axonal degeneration.

Direct investigations of blood flow at sites of nerve crush or transection

have not identified ischemia, even immediately after injury, or later as edema

develops [790,793]. Instead, blood flow is well maintained and rises to produce

hyperemia (high local blood flow). At crush sites, endoneurial nerve blood flow

begins to rise within 3 hours of injury, and further increases by 24 hours to peak

at 48 hours following injury. Partial nerve ligation, or complete nerve transection

renders similar changes. Following transection, hyperemia becomes particularly

intense within the proximal nerve stump but also occurs in the distal nerve

stump [281]. Hyperemia develops within both the endoneurium, demonstrated

by HC and in the epineurial vascular plexus observed using LDF studies. Finally,

early hyperemia demonstrated using physiological techniques also correlates

with morphological evidence, indicating vasodilatation of epineurial and endo-

neurial microvessels. Given all of these findings, it is not suprising that, in a set

of experimental studies, application of hyperbaric oxygen to treat presumed

hypoxia failed to ameliorate damage after nerve injury [246].

Early after injury, CGRP and nitric oxide (NO) associated with axon endbulbs

contribute toward nerve hyperemia. Other mechanisms maintain elevated nerve

blood flow at later time points. By 7–14 days after injury, for example, blood flow

measured in both the epineurial and endoneurial compartment persist above

normal levels. By this time, large rises in the numbers of vessels, particularly in

the epineurium, can be identified and are found both in the proximal and distal

nerve stumps (Figure 7.6). Moreover, accompanying this robust angiogenesis

are local rises in VEGF mRNA expression [281]. These features of peripheral nerve

repair parallel the changes of general wound healing.

It is also surprising that even long zones of nerve crush (e.g., 20mm in length)

do not develop ischemia in experimental studies (Xu, Zochodne et al., submitted

data). In this type of injury, it might be expected that vascular disruption and

edema would be particularly intense. These injuries not only retained normal or

increased levels of local blood flow but also supported robust regenerative

activity.

Despite these comments, there are instances of chronic compression or acute

compression where ischemia of the nerve trunk is likely responsible for axon

damage. These include large hematomas or mass lesions (or compartment

syndromes discussed above) that destroy significant portions of the vascular

plexus. None of the lesions or injuries discussed above may accurately reflect

the complex microenvironment of actual clinical lesions. Local edema and


vascular interruption could have an impact on multiple sources of vascular

supply to the nerve trunk. Extravasation of blood (and hemoglobin) along long

segments of the nerve trunk can quench NOmediated vasodilation. After stretch,

also a common clinical lesion, greater disruption of vessels may occur. While

quantitative studies of nerve blood flow have not been examined after stretch,

preliminary work using luminal dye studies suggests it can disrupt the vasa

nervorum [420].

Figure 7.6 Photomicrographs of peripheral nerves perfused with India ink to outline

their blood vessels as black profiles. The top panel shows a normal sciatic nerve of

a rat and the bottom image is from a proximal neuroma stump of a transected nerve

at 14 days. Note the very extensive investment of new blood vessels indicating

angiogenesis after injury, especially in the epineurial space. The images are unfixed

transverse sections of the sciatic nerve harvested after perfusion under anesthesia.

(Bar ¼ 200 microns) (Reproduced with permission from [805].)

167Local blood flow and the role of ischemia in acute nerve injury

Blood flow following long-standing nerve injuries

There are changes in blood flow of peripheral nerves that have sustained

chronic injuries. Distal nerve stumps that have not been reinnervated over

several months become less receptive to later reinnervation by axons, a problem

that is discussed further in Chapter 8. One of the factors that contribute to this

unreceptive microenvironment may be chronic declines in blood flow.

Changes in blood flow after chronic nerve injuries have been examined in

several models. Local blood flow was examined in 6-month neuromas formed

at the proximal stump of transected rat sciatic peripheral nerves [755]. Histo-

logically, these were complex lesions, with some zones retaining prominent

vascular investment, but nearby zones devoid of both blood vessels and axons.

Similarly, quantitative measures of flow within neuromas were highly variable

with hyperemia in some areas, but relative ischemia in others. Minifascicles

appeared in well-vascularized portions. It is possible that axons prefer to grow

in well-perfused areas of neuromas, or alternatively entrain new blood vessels

to accompany them.

Long-term experimental neuromas exhibit features observed in human

neuroma specimens. These kinds of chronic lesions were illustrated in a classical

pathological monograph by Lyons and Woodhall [431] depicting nerve segments

removed from soldiers injured in warfare. Resected several months after injury

during attempts at nerve repair, the specimens highlight zones of fibrosis,

atrophy of whole fascicles, and disrupted nerve trunk architecture. Where specifi-

cally examined, there were also instances of prominent thrombosis involving

nerve blood vessels, suggesting chronic ischemia.

Chronically denervated distal nerve stumps also develop ischemia. As in

neuromas, declines in their local blood flow may present an unsuitable micro-

environment for regeneration. In experimental studies of distal nerve stumps

that were not reinnervated, there were substantial declines in local endoneurial

and epineurial blood flow 3–6 months after injury [281]. The number and caliber

of epineurial microvessels declined at these later time points and levels of

mRNA for VEGF reverted to baseline values. In contrast, nerve trunks that had

undergone immediate resuture had preserved endoneurial nutritive blood flow.

Resutured nerves however, did exhibit persistent abnormalities of blood flow in

the epineurial plexus. Progressive fibrosis and scarring in the epineurium after

injury may permanently remodel its vascular plexus: it is uncertain whether

this change renders repaired nerve trunks more vulnerable to later ischemic or

mechanical injury.

Vascular changes in long-term damaged nerve trunks may also be secondary

to chronic loss of axons rather than the cause of poor regeneration. It is possible


that long-term declines in blood flow simply reflect a microenvironment with

lower metabolic demands. Alternatively, molecular connections among axons,

SCs, and endothelial cells of the vasa nervorum could also influence how severely

disrupted nerves are eventually supplied with microvessels. These may include

shared trophic factors, such as VEGF, or basement membrane components.

Summary

Peripheral nerve trunks and ganglia are supplied by blood vessels

termed vasa nervorum that possess unique physiological and structural qual-

ities. These qualities have an important impact on the response of the nerve

trunk to injury. Fortunately, the vascular supply of peripheral nerves is rich,

redundant, and difficult to render ischemic. Most focal injuries of nerve likely do

not involve ischemia Nonetheless, ischemic injuries to nerve from complete

interruption of the blood supply of a peripheral nerve can develop in some

situations and render acute axonal conduction block across the injury zone.

If prolonged, distal axonal degeneration ensues. Chronic nerve trunk lesions

are associated with substantial declines in local blood flow that may contribute

to an unfavorable microenvironment for regeneration.

Suggested reading

Zochodne, D. W. (2002). Nerve and ganglion blood flow in diabetes: an appraisal.

In Neurobiology of Diabetic Neuropathy, ed. D. R. Tomlinson. pp. 161–202. San Diego:

Academic Press [783].

Zochodne, D. W. & Ho, L. T. (1991). Unique microvascular characteristics of the

dorsal root ganglion in the rat. Brain Research, 559, 89–93 [792].

169Summary

8

Delayed reinnervation

What kind of reception do regenerating axons or SCs encounter when

they enter nerve trunks that have not housed axons for substantial periods

of time? Are target organs receptive to reactivation after months or years

of denervation? Unfortunately, the structural and molecular consequences of

prolonged denervation substantially diminish the likelihood for reinnervation.

This chapter deals with delayed reinnervation, a common and all too frequently

unavoidable problem in patients.

Clinical scenarios and long-term denervation

There are several reasons why axons may encounter denervated distal

stumps or target organs months or years after an injury. The first is obvious. The

most optimistic rates of axon recovery range between 1 and 3mm/day or an inch

per month. Many severe human nerve trunk injuries occur in large proximal

nerves, such as the sciatic nerve in the thigh or buttock [557] (see Figure 1.2).

These lesions rarely allow successful recovery of sensation or motor connections

to muscle endplates in the distal leg or foot. In the case of a lesion of the sciatic

nerve at the level of the thigh, it would require over a year for axons to regener-

ate an approximate distance of 800mm to the foot unimpeded. Moreover, distal

portions of the nerve trunk would not receive new axons for many months.

This analysis, however, oversimplifies the true scenario that may exist. Estimates

of regeneration are based on the time the first axons meet their target, whereas a

substantial population of axons must connect to their targets for functional

reinnervation. While this proportion might not have to match the original

complement of axons, sufficient number must be available and must be capable

of sending collateral sprouts to target organs in the muscle or skin. We now

170

know, however, that axons do not regenerate in a single wave or front. Rather,

early pioneer axons are followed by axons that advance at staggered rates. In a

crush injury, it is usually assumed that the “pathways” or Bands of Bungner

suitable to support regenerating axons are intact and allow rapid re-entry of

axons into a distal stump. This injury is associated with better regeneration, but

nonetheless only occurs along very protracted timelines. Simple crush injuries,

however, are only a small proportion of clinical injuries. Higher grade nerve

lesions with superimposed ischemia or added stretch impose new and added

constraints to axon regeneration.

Many injuries are not strictly localized to one site but may have separate

insults added in tandem. One of the most vivid examples of this seen by the

author was a victim of a bear attack and mauling while cycling in the Canadian

Rockies. The animal bit through and resected tissue that spanned much of the

victim’s arm from shoulder to elbow. Some of the nerve trunks that normally

course through this territory (ulnar and radial nerve) were completely destroyed

rendering denervated nerves and targets in the forearm and hand. The median

nerve, however, was found to be grossly connected and preserved in the bed

of severely damaged, swollen and resected muscle, skin, and bone. The combi-

nation of lesions involving stretch, ischemia, or crush along several centimeters

of its length can only be estimated. Surgical action to graft nerves and bridge

such lesions is challenging. In an acute situation with prominent tissue swelling,

a high risk of infection, and other problems, the probability of a successful result

is greatly diminished.

The second major reason that axons encounter long-term denervated territor-

ies is dictated by clinical circumstances. In the case of blunt nerve trunk injuries,

it is not recommended that the injury site be immediately opened, exposed,

and the integrity of the underlying nerves examined. Such an approach would

add to further traumatic damage already present in and around the nerve trunk.

Yet this type of early exposure, with electrophysiological recordings around

the injury zone might be essential to determine if the lesion is neurapraxic,

and likely to recover well without intervention, or has axon interruption,

i.e., axonotmesis. In the case of a neurapraxic lesion, the finding of conduction

block with retained excitability and function of the distal nerve trunk indicates

that full recovery can ensue. If the lesion is accessible by non-invasive electro-

physiological studies (e.g., nerves beyond the mid-upper arm and in the forearm

or leg below the knee), then conduction block can be identified early after the

injury. Thus, the identification of a predominantly neurapraxic lesion with

conduction block is important in directing surgeons not to intervene, since

spontaneous recovery is expected. In the case of aoxonotmesis, however, the

choice is less obvious. Some of these lesions, like the simple crush above, may

171Clinical scenarios and long-term denervation

recover spontaneously with time, particularly if the distance to the target is

short. Unfortunately, simpler approaches, including electrophysiological ones,

to determine which axonotmetic lesions are likely to regenerate enough for

a functional benefit are unavailable. Both the passage of time and careful

follow-up are required to determine if an axonotmetic lesion is beginning to

reinnervate distal muscles. In the author’s experience, some distal targets may

be reinnervated by a single motor unit, without the benefit of additional recon-

nected motor units several years in the future. Therefore, these lesions may

require re-exploration, their associated neuromas resected and a decision made

to either transect and resuture “refreshed” ends or graft the lesion. The decision

to resect and repair is therefore difficult when a lesion might recover better

without intervention. If further transection and resuture is required, the time-

table for regeneration begins anew.

A third major reason that axons encounter long-term denervated territories is

because they are common after transection injuries or neurotmesis. These injur-

ies impose additional barriers to regeneration and reinnervation of distal stumps.

After nerve trunk transection and immediate resuture, it is estimated that only

10% of axons reach their target, despite high-quality reconnection [744]. Moreover,

regeneration rates following transection are slower than the rates discussed

above. In a swollen and bleeding acutely injured limb, it may be very difficult to

identify sundered ends of peripheral nerves and to reconnect them microscopic-

ally. In the past, attempted high-quality reconnections of nerves have been post-

poned for 3 months or more to provide optimal conditions for surgical success.

In conclusion, all of these scenarios are common and approaches to them do

differ. The reader is referred to several excellent texts published on this topic

[349,421,660].

Proximal nerves and perikarya following axotomy

The proximal nerve stump and the pool of neurons supplying peripheral

nerves undergo several changes if a nerve trunk injury is chronic and does not

regenerate. With time, there may be further retrograde loss of parent neurons.

For example, while it is debated, axotomized motor neuron numbers may

decline by 60% over long periods [179]. This loss, in turn, reduces parent axons

available to send sprouts into the distal stump. RAG expression also gradually

declines in this situation; a proximal axon will not exhibit as robust an out-

growth capability as it might shortly after an injury. Fortunately, collateral

sprouting can compensate for some degree of irreversible neuron loss.

Lago and Navarro’s report [370] that proximal nerve stump axons regener-

ating into an isolated nerve stump maintain their population of axons for long

172 Delayed reinnervation

periods is more encouraging. Transected sciatic nerve axons were allowed to

regenerate into sections of nerves not connected to target tissues and subse-

quently formed neuromas. The regenerating axons remained intact, did not

drop out, and maintained their expression of immunohistochemical markers

(ChAT, CGRP, GAP43/B50) for up to 9 months later. These findings indicate that

long-standing transected nerves, as might occur after amputation, retain the

capability of reconnecting to a distal target for many months. The authors

suggested that these intact proximal axons might potentially be interfaced with

prosthetic devices.

Although axons may persist in proximal stumps that are not reconnected,

Furey et al. [204] show that motor neurons are less likely to extend new axons

following prolonged axotomy. The experimental design differed from that of

Lago and Navarro above. Femoral nerves were transected and either implanted

into a blind tube to prevent regeneration or into a cut saphenous nerve without

a target. The capability of motor neurons to regrow into a fresh distal stump was

then tested after a second surgical procedure. Chronically axotomized motor

neurons were compared with motor neurons allowed to regrow toward their

targets (primary resuture of the transected femoral nerves). Their ability to

regrow was measured with retrograde tracers. It was found that motor neurons

deprived of contact with their targets were less likely to regrow at a later time.

The authors also emphasize that counts of regrowing axons may not accurately

reflect the population of regenerating motor neurons because up to ten axons

may grow from a single parent motor neuron.

Prolonged disconnection from target tissues is a common clinical scenario

that impairs later regeneration. Parent neurons may disappear from retrograde

loss or may simply fail to respond to a regenerative stimulus. While these

neurons may send out multiple daughter sprouts, their fidelity and success of

functional reconnection cannot compensate for loss of parents.

Long-term denervated nerve trunks

For the first 2 or more weeks after a nerve trunk injury, active distal

Wallerian-like degeneration ensues. SCs proliferate, migrate, change their pheno-

type, and elaborate growth factors. There is an influx of macrophages, a local rise

in blood flow, and angiogenesis. These events are reviewed in previous chapters.

Overall, the behavior of cells and the synthesis of cytokines indicate an active

inflammatory milieu.

When reinnervation is prevented, these features dissipate over time. The

microenvironment of denervated distal trunks eventually becomes unsupportive

of regeneration [656]. Even after 2 months of delayed repair, 20% fewer motor

173Long-term denervated nerve trunks

neurons regenerated axons in the femoral nerve [658]. SC proliferation tapers,

and they elaborate lower levels of growth factors and cytokines. With further

time, the number of SCs decline, they lose their reactive phenotype, and become

atrophic [179]. SC molecular markers associated with an activated status

decrease: erbB2, erbB4, and p75 [397,768]. While SCs can later be partly reacti-

vated, they remain less supportive for axons seeking to regrow into the distal

nerve stump. Associated basement membranes and Bands of Bungner eventually

disappear in denervated distal stumps [215]. Interestingly, while smaller

numbers of atrophic SCs cannot adequately support new axons, they are capable

of remyelinating those axons that do eventually reach them [656]. SCs are also

conditioned by the types of axons that regrow within them and thereafter choose

suitable axons. For example, Sulaiman and colleagues [658] found that motor

axons will enter and reinnervate a graft that originated from a sensory nerve

trunk. If, however, sensory axons are simultaneously allowed to reinnervate

it (i.e., into the other end of the graft), motor axon ingrowth is substantially

reduced. The sensory axons appear to sensitize the residual SCs to support the

original axons they once associated with. In this case, sensory axons rapidly

instructed the nerve graft to recapitulate the allegiance to sensory, not motor

axons. These studies show that both chronic denervation of the distal stump and

also “sensitizing” the stump with sensory axons substantially impairs motor

regrowth. It seems likely that regrowing motor and sensory axons from mixed

nerves compete for SC support.

Declines in a number of cell types have been described in chronically dener-

vated nerve stumps. For example, the numbers of overall proliferating cells fell

between 3 and 6 months in chronically denervated distal nerve stumps [626].

Macrophage numbers also declined. In studies of chronically denervated rat

sciatic nerves lasting 3–6 months, blood flow within both the endoneurial and

epineurial compartments fell below levels of intact nerves [281]. When chronic-

ally denervated nerves are examined grossly, they appear atrophic, fragile, and

translucent. There is atrophy of endoneurial fascicles and collagen replaces

considerable portions of the nerve architecture (Figures 8.1, 8.2). Long-term

denervated nerve stumps become “fibrotic” because of collagen deposition; the

epineurium and perineurium abundantly express Types I and III collagen mRNA

in denervated stumps. Axon re-entry into chronically denervated distal stumps is

associated with a concurrent increase in mRNA of Type I collagen [626].

Chronic denervation is associated with reductions in the the mRNAs of

growth factors synthesized by SCs, mast cells, and macrophages. For example,

GDNF levels fall progressively (after a large rise soon after injury) in the SCs

of distal nerve stumps up to 6 months if they are not reinnervated [278].

Hall showed that chronically denervated SCs had lower proliferation rates,


less migratory potential, and attenuated expression of cell adhesion molecules

(N-CAM and N-cadherin). Some of these features could be restored with neuregulin

(GGF) [248]. VEGF mRNA, a transcript that rises to high levels with angiogenesis

early after injury, declines to low level baseline values during prolonged dener-

vation [281]. Finally, chronically denervated nerve stumps may impose additional

barriers to regeneration through the expression of chrondroitin sulphate pro-

teoglycans (CSPGs). CSPGs collapse growth cones and inhibit nerve regeneration.

Neural stem stells that expressed MMP-2 (matrix metalloproteinase-2) facilitated

regrowth into chronically denervated nerve stumps by degrading CSPGs [258].

Denervated muscle

Muscles deprived of their innervation undergo substantial alterations.

These include structural changes in muscle fibers, as well as alterations in

electrophysiological and biochemical properties. Most are reversible during

(a) (b)

2

Figure 8.1 Images of peripheral nerve specimens from WWII soldiers wounded in

battle by missile injuries that were resected during attempts at surgical repair several

months later. The example in A is a longitudinal section of a sciatic nerve illustrating

severe disruption of the nerve architecture and a neuroma in continuity (white arrow).

From a similarly injured sciatic nerve, the image in B is from the nerve distal to an

injury (the insets are proximal, left, and distal, right, boxed, to the injury) where no

significant axonal regrowth has developed. Note the fascicles are small and atrophic

and it is likely that they consist mainly of fibrotic collagen. These distal stumps do not

provide a favorable microenvironment for regrowth. (Reproduced with permission

from [431].)

175Denervated muscle

reinnervation. Their onset is more rapid when the nerve injury is closer to the

muscle. There is debate over whether denervation changes occur as a result of

loss of activity, loss of input by motor nerves, or from loss of axon trophic

molecules (see review [471]). Experimentally, rigorously applied forms of disuse

can recapitulate several features of denervated muscle. Many changes of dener-

vation can be reversed by direct stimulation of muscles, a traditional clinical

approach for “conditioning” muscles during recovery from a nerve injury. In the

case of the neuromuscular junction, however, there is evidence that specific

molecules are involved in the remodeling during reinnervation. These include

Figure 8.2 Examples of isolated minifascicles (arrows) from a chronic (6 month)

experimental neuroma from a rat sciatic nerve. Much of the field is not invested

by axons or blood vessels (asterisk). The image is from a semithin section, embedded in

epon and stained with toluidine blue.


IGF-II, agrin, and neuregulin. CNTF may act as a “myotrophin” in denervated

muscles, capable of attenuating atrophy and reducing loss of twitch and tetanic

tensions associated with denervation [261].

The weight and force of muscles decline to 30% of their original value after

6 months of chronic denervation [199]. In the author’s experience, dramatic

atrophy of human muscles can occur within a similar time frame during dener-

vation. The changes in muscle, however, do not indicate resistance to reinnerva-

tion. Failure of motor axons to return to chronically denervated endplates is

the predominant reason for poor motor reinnervation. Intrinsic changes in the

target muscles such as fibrosis and calcification occur much later. Recovering

and reconnecting motor axons can also enlarge their motor units by five times

or more [199].

Denervated muscles thus develop atrophy of individual muscle fibers and

gross loss of muscle weight. Moreover, these changes may be accelerated in slow

twitch muscles. By light microscopy, atrophic denervated muscle fibers form

“nuclear knots” consisting of chains of nuclei occupying very little cytoplasm.

Ultrastructural changes in muscle may occur within days of denervation. There

is disorganization of sarcomeres and disruption of myofibrils. Sarcoplasmic

reticulum changes include hypertrophy of terminal cisternae, proliferation of

transverse tubules, and rises in triad numbers [471]. Eventually, there may be

frank loss of muscle fibers. The gross weight of muscles is not, however, an

accurate index of denervation or reinnervation. Its use in experimental work

should be discouraged; replacement of muscle fibers by connective tissue and

other technical factors may independently alter weight measures. Morphological

features of reinnervated muscles, such as fiber caliber, provide more accurate

estimates of denervation and reinnervation. Other late morphological changes

of denervated muscle are vacuolar degeneration, hyaline degeneration, and

progressive disintegration of fibers [471]. Electromyography of long-term dener-

vated muscles in humans (10–20 years or more) identifies electrical silence and

a “woody”or fibrotic gross consistency. This state is severe and likely at this time

to be an irreversible state of affairs. Reinnervation is then essentially impossible.

The behavior of terminal SCs, discussed in detail in Chapter 6, is important in

promoting appropriate reinnervation. Since terminal SCs guide axons back to

denervated endplates, their loss may result in failed axon pathfinding. This has

been identified (see review [319]) in situations where denervation is prolonged.

The plasticity of terminal SCs, so apparent during early reinnervation or partial

denervation with concurrent reinnervation, may gradually dissipate. Endplates

may therefore be abandoned. Regenerating motor axons, strictly guided by

terminal SCs are unable to identify these abandoned sites, and they are not

reinnervated [319].

177Denervated muscle

Electrophysiological recordings, described in Chapter 4, can also provide infor-

mation about delayed or frustrated reinnervation. CMAPs decline in amplitude

in direct proportion to the loss or degree of dysfunction of the motor axons

innervating the endplates being recorded from. Recovery of CMAPs similarly

reflects the extent of axon reconnection, but collateral sprouting to other

denervated muscle endplates also occurs at this time. Rises in CMAPs during

reinnervation therefore reflect both reconnection of regenerated axons as well

as collateral sprouting (see Chapter 6). Serial studies with persistent small

amplitude CMAPs or NAPs that do not recover or change with time indicate that

further regeneration has halted.

Needle electromyography may provide further information. With long-term

denervation, some have argued that the amplitude of fibrillation potentials

gradually declines, reflecting individual muscle fiber atrophy. Muscles with very

prolonged denervation fail to generate an injury discharge with movement of

the needle electromyography electrode, a property of normal muscle.

Contractile changes develop within denervated muscles. These include the

acquisition of fast contractile properties by slow twitch muscles whereas fast

twitch muscles are not altered. Overall, contraction in both fiber types is slowed,

and tension development rate and velocity are reduced [471]. Denervated

muscles have decreased uptake of glucose, impaired binding of insulin, deple-

tion of glycogen, altered activity of glycolytic enzymes, and decreased glucose

oxidation [81,147,151,275,312,494]. Relatively few studies of denervated muscles

in humans are available since even rapid freezing of biopsy specimens can alter

their biochemistry. A noninvasive approach is to analyze muscle bioenergetics

using NMR (nuclear magnetic resonance) spectroscopy, a technique that meas-

ures levels of phosphocreatine (PCr), inorganic phosphate (Pi), ATP, and ADP.

Denervated forearm muscles in patients with nerve injuries or motor neuron

disease were studied by P31 NMR spectroscopy. Subjects placed their arms into

a small bore (26cm bore, 1.89T) magnet [810]. Patients were compared with

normal controls and to patients with forearm disuse resulting from wearing

orthopedic casts for bony injuries. Denervated muscles had reduced ratios of

PCr/Pi, indicating a decline in the bioenergetic reserve of the muscle. Denervated

muscles also had a rise in pH. The caveat is that these changes are not specific

and other primary muscle disorders are capable of generating them. The time

course of denervation related changes was examined experimentally by Dort

et al. [148]. Declines in PCr began within 1 week of denervation of the rat facial

muscle and continued out to 8 weeks following injury.

Long-term denervation also results in contractures of muscle tendons that

reduce jointmobility andmay fix the limb in postures that interfere with function.

In the human foot, a shortened Achilles tendon leads to persistent plantar flexion


of the foot. In rodent models, persistent flexion–contraction of the paw digits may

develop. Functional tests of motor recovery are complicated by these changes [759].

Long-term denervated skin and other organs

There are long-term changes in other endorgans with prolonged

denervation. In patients with chronic polyneuropathies, skin atrophy and loss

of hair may closely parallel the extent of sensory loss. Denervated skin is also

prone to breakdown and ulceration. In part, this occurs because desensitized

skin may undergo unrecognized injuries. Patients may not sense the presence of

a nail or a stone in their shoe. Loss of sudomotor innervation also contributes to

ulcerations in denervated skin. Normal sweating likely moisturizes and protects

the skin. Skin deprived of sweating can become thickened, cracked, and suscep-

tible to injury and infection.

Sensory nerves also supply the skin and other organs with neuropeptides that

influence its microcirculation and healing. Two of the major neuropeptides

considered are SP (Substance P) and calcitonin gene-related peptide (CGRP).

Neurogenic inflammation refers to the participation of sensory nerve terminals

in the inflammatory response of the tissues they innervate. All of the principal

components of inflammation, described as the “triple response of Lewis,” are

involved: tumor (swelling from edema, plasma extravasation), rubror (redness

from rises in local blood flow), dolor (pain), and calor (heat from rises in blood

flow) [395]. The triple response refers to the timing of these changes, following

a firm stroking of the skin. Lewis’ classical description is that of an initial

dull red line followed by a bright red halo called the flare response, then

third, swelling (edema) with blanching of the line. When tissues are denervated,

they are deprived of neurogenic inflammation and the overall inflammatory

response is attenuated. Since all of the features of inflammation are important

in proper wound healing, denervated skin and other tissues have impaired

healing.

A number of properties of healing wounds are impaired during denervation

[32]. These include delayed contraction of their edges to close the wound,

reduced microvascular responses, delayed and prolonged inflammatory phases,

hypertrophic scarring, and delayed epithelial cell migration. Rises in blood

flow supply the demands of proliferating and activated cells and help with

clearance of debris and micro-organisms. Plasma extravasation delivers blood-

borne trophic and other molecules as well as macrophages to the injury site.

These properties, in turn, depend on neuropeptides released by neurogenic

inflammation. SP contributes to vasodilation and plasma extravasation. CGRP

is a very potent vasodilator [60]. Both peptides may signal mast cells to release

179Long-term denervated skin and other organs

histamine to promote further vasodilation and plasma extravasation. Mast

cells also release proteases and other growth factors within the injury milieu.

In the absence of “peptidergic” sensory axons that normally innervate the

skin, all of these properties of neuropeptides that are required for healing

are absent.

Overall then, denervated skin that fails to mount an appropriate inflamma-

tory response to an injury experiences prolonged, impaired healing. Chronic

ulceration with superimposed infections result. Bone infection or osteomyelitis,

in turn, may be associated with loss of distal limb viability, requiring drastic

measures such as amputation. In diabetic patients with chronic polyneuropathy,

amputation due to nonhealing infected ulcers and osteomyelitis is a serious and

frequent complication. Fluctuating discoloration, particularly with a dependent

limb is a common development. Altered vasomotor control due to denervation

probably accounts for these changes. For example, loss of adrenergic vasocon-

striction of skin blood vessels may result in its dark blue discoloration (e.g., of the

toes and feet in sensory neuropathy). Loss of the muscle supply to a limb may

also cause edema because of impaired venous return, a function of the normal

contracting muscles of a limb.

Terminal sensory axon SCs, described in Chapter 6, can survive and support

regrowth to correct targets after denervation [152]. With prolonged denervation,

however, these specialized SCs become atrophic and disappear. Munger described

persistent changes in several skin sensory organs including Pacinian corpuscles,

Meissner’s corpuscles, Merkel touch domes, and the piloneural complex associ-

ated with guard hairs after prolonged denervation [483].

Collateral sprouting, described in previous chapters, is an important mechan-

ism for reinnervating skin or muscle when their nerve supply is permanently

interrupted [152]. While nociceptive axons undergo brisk collateral sprouting,

large caliber axons supplying touch/pressure may not. The hindpaw is thought

to exhibit less collateral sprouting of cutaneous axons than the dorsal trunk

skin. The presence of fewer available intact neurons in adjacent territories also

diminishes the formation of collateral sprouts.

Summary

Both distal portions of the peripheral nerve trunk and target organs

may experience prolonged loss of their axon investment. This is a major clinical

problem following peripheral nerve injuries. First, repair is not always accom-

plished immediately and second, regrowing axons require time to extend long

distances. Long-term denervated muscles have altered structural, electrophysio-

logical, and biochemical properties. Prolonged denervated skin has impaired


healing in part related to loss of neurogenic inflammation. In both muscle

and skin, specialized terminal SCs are important sources of guidance during

regrowth but may disappear with prolonged denervation.

Suggested reading

Fu, S. Y. & Gordon, T. (1997). The cellular and molecular basis of peripheral nerve

regeneration. Molecular Neurobiology, 14 (1–2), 67–116 [200].

Gordon, T. & Fu, S. Y. (1997). Long-term response to nerve injury. Advances in Neurology,

72, 185–199 [227].

181Summary

9

Trophic factors and peripheral nerves

The concept that molecules can signal neurons and axons and convince

them to behave in new, even innovative ways is exciting. Trophic factors are

molecules, usually proteins, that act on specific cell receptors to induce changes

in protein synthesis, outgrowth, or survival. In the nervous system, nerve growth

factor (NGF) leads the classic family of growth factors known as “neurotrophins.”

A working definition of a neurotrophin is “an endogenous soluble protein

regulating survival, growth, morphologic plasticity or synthesis of proteins for

differentiated function of neurons” [256]. NGF was discovered in 1951 by Rita

Levi-Montalcini and subsequently characterized by Stanley Cohen, culminating

in the award of the Nobel Prize in Physiology or Medicine in 1986 to both investi-

gators [388,389]. Using sarcoma tumor in the mouse model, Levi-Montalcini and

Viktor Hamburger established that NGF was a soluble factor from mouse sarcoma

tumor that was capable of inducing hyperplasia in sympathetic ganglia [389,390]

(Figure 9.1). The original articles describing these discoveries are recommended

and make for fascinating reading [390]:

The growth promoting effect of the tumor is mediated by a diffusible

agent. This mode of action was suggested by the observation that

sympathetic ganglia which are located rostrally to the tumor and

not connected with the tumor by nerve fibers, are also conspicuously

enlarged. The overgrowth of the ganglia is the combined result of an

increase in cell number and in cell size. The enlarged ganglia send large

sympathetic fiber bundles into the adjacent viscera. Branchial and

abdominal viscera which normally receive only a sparse innervation

or none at all are inundated with fiber masses during the second and

third weeks of invasion.

182

Growth or trophic factors support neurons through several mechanisms

that include retrograde transport from target tissues, paracrine support from

other cell types adjacent to neurons, or autocrine support from themselves or

neighbor neurons. Growth factors can be elaborated by a variety of cell types

including neurons, SCs, perineuronal satellite cells, and other cells of target

organs.

The topic of growth factors and neurons is large and this discussion is limited

to its impact on the adult peripheral nervous system. The reader is also referred to

Boyd and Gordon’s [58] comprehensive review of this topic. Multiple and overlap-

ping roles exist for growth factors during development both in the peripheral

and central nervous systems, and in synaptic plasticity. These important roles

for growth factors are not addressed here.

Chickembryonic

tissue

Fragment ofmouse sarcoma

Figure 9.1 Illustrations of the original findings of Levi-Monalcini and colleagues

that led to the discovery of NGF. Axons emerging from explanted sympathetic

ganglia grow toward mouse sarcoma cells that secrete a gradient of NGF. (Based on

the original figures by Levi-Montalcini [389] and reproduced with permission

from [782].) See color plate section.

183Trophic factors and peripheral nerves

The neurotrophin family

Since the publication of the original NGF papers, the neurotrophin

family has expanded to include several new members all sharing an interaction

with the p75 neurotrophin receptor (previously known as the “low affinity” NGF

receptor) but operating on more specific Trk (“tropomyosin-related kinase”; tyro-

sine kinase) receptors. They are brain-derived neurotrophic factor (BDNF) [31,386],

neurotrophin-3 (NT-3) [276,436], and neurotrophin 4/5 (NT-4/5) [42]. Neurotrophin-6

(NT-6) and neurotrophin-7 (NT-7) are regarded as additional members but NT-6 is

nonsoluble and both are only found in fish [230,371]. There is high sequence

homology (approximately 50%) among the neurotrophin members [58].

In the classical scheme, NGF binds to TrkA, BDNF and NT-4/5 to TrkB, and NT-3

to TrkC (Figure 9.2). Work by Verge and others, however, has established the

classical schema is incomplete. These interactions are actually more complex

and expression of Trk receptors can change following injury [231,324]. Some

neurotrophins can interact with other Trks, e.g., NT-3 with TrkB and TrkA.

Overall, Trk receptors mediate differentiation, survival, and loss of proliferative

capacity through their intracellular cascades. TrkA receptors have been typically

Figure 9.2 Illustration of members of the neurotrophin growth factor family and

their receptors. (Reproduced with permission from [782].) See color plate section.

184 Trophic factors and peripheral nerves

localized to small sensory neurons with unmyelinated axons that mediate

nociception and temperature sensation and contain Substance P (Figure 9.3).

TrkA are also found on NGF responsive postganglionic sympathetic neurons, as

originally described by Levi-Montalcini. TrkB receptors are located on medium-

sized sensory neurons and motor neurons, while TrkC receptors are located

on large caliber sensory neurons and motor neurons with large myelinated

axons [325,457]. As might be expected, Trk receptors are also found in the central

nervous system, and the distribution of the receptors (and the response of the

neurons bearing them!) vary dramatically during development. At different

stages of development, the neurotrophins and their receptors have specific

responsibilities. For example, NT-3 and TrkC are required for the development

of muscle spindle afferents and larger caliber sensory neurons. Many investi-

gators have argued that their importance is much greater during development

than in adulthood. For example, axotomy interrupts the retrograde transport

of NGF and other growth factors from target tissues and massive retrograde cell

death in neonatal sensory neurons occurs. In contrast, adult nerves subjected to

the same injury suffer relatively little retrograde apoptosis (see Chapter 3),

indicating that they have less reliance on ambient growth factor transport.

There is also extensive literature on neurotrophin family members and on their

modulation of synaptic transmission.

Sympathetic neurons

TrkA NGF

TrkA NGF

TrkB

TrkB

TrkC

BDNF,NT-4/5

NT-3

Nociceptive fibers

Mixed afferents

Motor neurons

Large diameter afferents

Figure 9.3 Illustration of the peripheral neuron subtypes responsive to members

of the neurotrophin growth factor family. (Reproduced with permission from [782].)


185The neurotrophin family

Neurotrophin growth factor receptors [325] are differentially distributed

among neuron populations. TrkA, high affinity NGF receptors, are found on

41% of lumbar DRG neurons. TrkB, the cognate receptor for the BDNF on 33%,

and TrkC, the receptor for NT-3 on 43%. The low affinity neurotrophin receptor,

p75, is capable of binding all neurotrophin family members (Figure 9.2) and

is found in 79% of DRG neurons. There are two TrkA isoforms that have overlap-

ping expression. This clearly establishes that extensive colocalization of Trk

expression exists. Of all sensory neurons, 10% coexpressed TrkA and TrkB, 19%

coexpressed TrkA and TrkC, and 18% coexpressed Trk B and Trk C. Trilocalization

of Trks was rare, but Trk and p75 colocalization was tightly coupled. Neurons

expressed only one Trk 34% of the time and more than one in 40% of lumbar DRG

neurons. Interestingly, approximately 26% of DRG neurons that were small and

medium in size did not express Trks and represented the IB4, or GDNF responsive

subpopulation. A small number of neurons and satellite cells also expressed

truncated isoforms of Trks (especially TrkB,C) that do not have an intracellular

signaling domain [579]. These splice variants may allow the receptors to have

more promiscuous binding to other, nonpreferred neurotrophins, but their

function is unknown. They may modulate interactions of the ligand with full-

length receptors (dominant negative). They may also alter rates of degradation

of ligand–receptor complexes or limit the overall concentrations of the neuro-

trophins available to act on normal Trk receptors [325].

Unlike other neurotrophins, BDNF is synthesized in intact DRG sensory

neurons and its levels rise within 1 day following injury [686]. Motor neurons

do not express BDNF. It is likely that all neurotropins, however, are internalized

by axons and are retrogradely transported from target tissues. BDNF undergoes

both anterograde and retrograde axoplasmic transport, whereas NGF only

exhibits retrograde transport [142,262,606,607]. The transport of NT-3 is contro-

versial, but it does undergo rises in local expression by target tissues (such as

muscle) after injury, possibly from local synthesis [686]. The classical tenet of the

“neurotrophic factor hypothesis” is that such target-derived support, delivered

by retrograde transport, is essential to the maintenance of neurons. During

development or regeneration, neurons compete for these sources of support,

and axons that fail to secure them undergo apoptosis (Figure 9.4).

Neurotrophin receptors and signaling cascades

Neurotrophins are capable of signaling neurons locally, for example,

at growth cones, or centrally at either the cell body membrane or within its

cytoplasm after transport [679]. The initial synthesis of NGF is in the form of a

proNGF dimer molecule. The proNGF chains are then cleaved to form b subunits,

and two g and a subunits are added to form a storage form of the molecule


[15,16,453,622]. Finally, each b subunit joins an identical subunit to form a

noncovalently linked dimer [782] that is freed from its subunits and released

as the active form of NGF. Each b subunit is a necklace of 118 amino acids

connected by three covalent sulphur bonds on cysteine residues. The three-

dimensional structure of each NGF b subunit includes three antiparallel strands

that form a surface where they can therefore associate into a dimer, facilitated

by a cysteine knot motif. The antiparallel strands within each subunit are joined

by four loops that are unique among neurotrophin family members.

The mature NGF dimer is distinguished as having the attributes of a classic

signaling molecule. More recent evidence, however, has also indicated that its

precursor proNGF can act as a signaling molecule with a repertoire of actions

different from its cleaved substrate [178]. Its overall levels may therefore be

determined by the efficiency of proteolytic conversion to mature forms. ProNGF

binds p75 receptors with higher affinity and may, as a consequence, have more

widespread actions than classical NGF [293,383]. Several forms of proNGF have

also been described [178].

Mature or classical NGF binding to TrkA results in receptor dimerization and

autophosphorylation of its intracellular domains. Trks are concentrated in

caveoli-like structures that facilitate their interaction with proteins on lipid rafts

[291]. Ligand binding stimulates their internalization through clathrin-coated

pits and macropinocytosis [579].

Other members of the neurotrophin family also operate as noncovalently

linked homodimers. Like NGF, their ligation of Trk receptors causes receptor

dimerization, then autophosphorylation of an intracellular “activation loop” of

the receptor. Specific phosphorylation sites (called Y490 in the case of TrkA)

Axoninitiation

NGF

Incorrecttarget

Correcttarget

Correcttarget

Absent orremovedtarget

Contact withtarget tissues

Death ofincorrectlyprojectingneurons

Figure 9.4 Illustration of the neurotrophic hypothesis. Neurons from axons that arrive

at a target source of growth factor survive whereas those that fail to arrive or grow

to incorrect targets undergo apoptotic cell death. (Reproducedwithpermission from [782].)


187Neurotrophin receptors and signaling cascades

interact with proteins containing SH2 (src-homology-2), that in turn act as

scaffolds to activate one of four pathways: PI3-kinase, PLC-g, and Ras. Another,

but less well-studied pathway involves SNT [suc1-associated neurotrophic factor-

induced tyrosine phosphorylated target] [58]). Some of the protein intermediaries

include shc, Gab-1, Grb2, and Sos (for review see [58]).

All of the mature neurotrophins (as well as proNGF) thus interact with p75,

a member of the tumor necrosis receptor superfamily (Figure 9.2). P75 is a

multifunctional molecule, facilitating neurotrophin signaling but also capable

of promoting cell death. It interacts with its neurotrophin ligand and with

the Trk receptors in complex patterns (see review [579]). Thus, p75 may inhibit

activation of Trks by nonpreferred neurotrophins, thereby enhancing the speci-

ficity of neurotrophin–Trk interactions. It may also potentiate activation of Trks

when there are low levels of ligand, and it may promote retrograde transport of

neurotrophins. Overall, while the interaction of Trk and p75 therefore facilitates

neurotrophin signaling, whether this requires a direct physical interaction to do

so is also debated [735].

In the absence of its Trk interaction, p75 also promotes apoptosis and cell

death through an intracellular “death” domain and it facilitates RHO GTPase-

mediated growth cone collapse (see Chapter 5). Proneurotrophins, such as

proNGF, bind with high affinity to p75, but they do not activate Trk receptors.

Other proteins interact with p75 and Trks. For example, Necdin is a multifunc-

tional signaling protein that facilitates the interaction between TrkA and p75 for

sensory neuron survival signaling [366].

The PI3K (phosphatidylinositol 3-kinase) pathway, introduced in Chapter 5, is one

of the main downstream targets of neurotrophin signaling (Figure 9.5). PI3K

interacts with protein complexes that form in association with the cytoplasmic

side of the cell surface. PI3K converts PI(2)P to PI(3)P (phosphatidylinositol

trisphosphate) that, in turn, phosphorylates and activates the serine/theonine

kinase Akt (also known as PKB). Akt then acts through a variety of pathways to

influence cell survival and outgrowth including the inhibition of GSK3b (that, in

turn, inhibits growth cone advance). Akt also activates IkB promoting its degrad-

ation and dissociation from NF-kB, another important player in sensory neuron

survival [182,249]. A novel pathway involves an NGF-p75 interaction, independent

of TrkA, that phosphorylates and inhibits phosphatase and tensin homolog

deleted on chromosome 10 (PTEN) [22]. PTEN, also introduced in Chapter 5, is a

phosphatase that inhibits PI(3)P formation and pAkt to thereby facilitate GSKbB

signaling. By “knocking down” PTEN, elevating pAkt, and thereby inhibiting

GSKb signaling, NGF-p75 can facilitate outgrowth independent of TrkA. The

end result of these many steps is an elevated level of pAkt. Thus either TrkA

signaling, or the p75 pathway, may increase pAkt.


A related growth factor signaling pathway involving PLC-g1 operates through

inositol 1,4,5-triphosphate (IP3) and diacylglycerol (DAG) to mobilize Ca2þ and

PKC. While its impact is uncertain, its actions on PKC may also be important in a

number of additional pathways including localization of GAP43/B50 to the

membrane of the growth cone.

The third major pathway of neurotrophin signaling involves the extracellular

signal-regulated protein kinases (ERKs), a subfamily of the mitogen-activated protein kinase

(MAPK) cascade. Growth factor receptors, as discussed above, interact with Ras. Ras

then activates the protein kinase Raf, that in turn phophorylates MEK1 or MEK2

(MEK ¼ MAPK and ERK kinases) to phosphorylate and activate ERK1 and ERK2.

This pathway eventually activates the intermediate early genes fos and jun. ERK5

is an additional target of neurotropin signaling and there are likely other related

intermediaries with varying evidence for their involvement [321]. ERKs influence

transcription within nuclei through several additional molecules, in particular,

cAMP response element binding protein (CREB) [579]. Neurotrophins elevate

cAMP levels, in turn, activating a series of pathways that support neurite out-

growth and involve PKA and CREB. All of these findings therefore illustrate that

CREB is a major mediator of neurotrophin actions [670].

There are a number of other proteins and signaling cascades that are

either influenced by neurotrophins or work together with them. Poly ADP-ribose

polymerase (PARP) acts on histone H1 of the chromatin structure to allow

RTK

PIP2

PIP3

PTEN

Akt

Activated

Pl-3 Kinase

GSk3

GSk3

PDK1 mTOR

Akt

Forkhead

Forkhead Survival

Survival

IKKIKB

NF - kB

NuclearImport

IKB

p50 p50

p50

p65 p65

p65

FastTranscription

Caspaseactivator

Apoptosis

IKK

Figure 9.5 Illustration of the PI3K-Akt neuron survival pathway activated by growth

factor receptors containing a receptor tyrosine kinase (RTK) intracellular signaling

domain. (Illustration by Scott Rogers based on diagrams posted by EMD/Calbiochem

(www.emdbiosciences.com).) See color plate section.


neurotrophin signals access to DNA. This access can thereby influence DNA

transcription and repair [711]. Neurotrophins activate RAC and CDC42 GTPases,

growth cone molecules that facilitate neurite outgrowth through PI3K. More-

over, NGF is capable of bypassing inhibition by CSPGs, an inhibitor pathway that

acts through RHOA GTPase. Trk receptors can be phosphorylated without neuro-

tropin ligation. For example, adenosine or PACAP signals through G-protein-

coupled receptors to activate Trk [570]. NGF signals converge with those mediated

by preconditioning and laminin–integrin signaling on GSK3b–APC [780]. Further

pathways that may operate independently of neurotrophins upstream but

eventually converge on the same transcriptional pathways, such as CREB, in-

clude JAK-STAT, Bcl-2, and PKC (see review [670]).

Other growth factors and the peripheral nervous system

The list of trophic molecules that influence peripheral neurons con-

tinues to expand. While many are proteins, others like purines, nucleosides,

and nucleotides [283,502] are not. The term “pleiotrophic” has been applied to

this broad group, highlighting their widespread actions beyond the nervous

system. This is not to be confused with a specific growth factor pleiotrophin

(PTN) described below. Alternatively, many growth factors originally linked to

other tissues, such as epidermal growth factor (EGF) or erythropoietin (EPO),

were later discovered to also act on the nervous system. A current list of several

of these growth factors is given in Table 9.1.

One class of growth factors, also called neurotrophic or neuropoeitic cyto-

kines, share a common receptor subunit known as gp130. The association of

this subunit with other specific receptor subunits confers ligand specificity. This

family also shares a single, four-a-helix bundle protein structure that, unlike

neurotrophins, does not form dimers [259,260]. The ligand–receptor complexes

signal through the cytoplasmic tyrosine kinase Janus-kinase (JAK) and signal

transducer and activator of transcription (STAT) pathways. The STAT3 transcrip-

tion factor promotes neurite outgrowth. The suppressor of cytokine signaling

(SOCS3), a member of another large protein family, dampens STAT3 actions

through feedback inhibition [469].

Members of the neuropoeitic cytokines include ciliary neurotrophic family

(CNTF), a growth factor that supports the survival of parasympathetic, motor,

sensory, and sympathethic neurons [299]. CNTF can rescue embryonic and

neonatal motor neuron cell loss after axotomy but the actions in adults are less

robust [232,398,537,593]. Its receptor complex consists of gp130, LIFRb, and

CNTFRa. Other members of this family include leukemia inhibitor factor (LIF; recep-

tors gp130 and LIFRb), interleukin-6 (IL-6; receptors gp130 (2 subunits) and IL6Ra),

and cardiotrophin-1 (CT1; receptors gp130 and LIFRb). CT1 is a survival factor for


developing motor neurons, but not adult motor neurons [538,593]. Oncostatin-M

(OSM) and neurotrophin-1/B-cell stimulating factor-3 are additional members

[259,614]. The roles of IL-6 and LIF in SC support, inflammatory signaling, and

regeneration are discussed in Chapters 3 and 5.

Table 9.1 Neurotrophin and non-neurotrophin growth factors

Name Receptor Peripheral neuronal target

Neurotrophin growth factors

NGF TrkA, p75 Small sensory, sympathetic neurons

BDNF TrkB, p75 Sensory neurons, motor neurons

NT-3 TrkC, p75 Large sensory neurons

NT-4/5 TrkB, p75

Non-neurotrophin growth factors

1. Acting through gp 130 receptor complex

CNTF LIFRb, CNTFRa Motor neurons, probably others

LIF LIFRb ?Several types

CT-1 LIFRb ?Several types

IL-6 IL6Ra ?Several types

2. Growth factors discovered in other tissues

FGFs FGFR1–4 ?Several types

EGF EGFR ?Several types

TGFb TGFbR-I-III ?Several types

BMPs BMPRII, ALK2,3,6 ?Several types

PDGF PDGFRa and b ?Several types

HGF Met TK ?Several types

MSP c-Ron ?Several types

VEGF Flk-1 ?Several types

EPO EpoR Sensory neurons, ?others

3. Insulin-related growth factors

Insulin IR Motor, sensory, ?sympathetic

IGF-1 IGF-IR Motor, sensory, ?sympathetic

IGF-II IGFII-R ?similar to IGF-1

4. Other cytokines

IL-1 IL-1R1, 1R2, 1RAcP ?Several types

?IL-11 ?Several types

5. Glial-derived neurotrophic factors

GDNF GFRa-1, c-RET Motor neurons

Neurturin GFRa-2, c-RET

Artemin GFRa-3, c-RET Sympathetic, sensory neurons

Persephin GFRa-4, c-RET

6. Other growth factors

PTN LRP-1, others ?Several types


Glial-derived neurotrophic factors (GDNFs) include GDNF, neurturin, artemin, and

persephin, members that share approximately 40% homology [30]. GDNF was

first described as a potent surval factor for midbrain dopamine cells [406,407].

Like neurotrophins, GDNFs have a cysteine knot motif that facilitates their

formation into homodimers. GDNFs, however, bind to a unique receptor com-

plex consisting of a series of GFRa receptors (GFRa-1 for GDNF, GFRa-2 for

neurturin, GFRa-3 for artemin, and GFRa-4 for persephin) and c-RET, a tyrosine

kinase [597]. When the two receptors and ligand interact, downstream signaling

ensues through recruitment of RET into lipid rafts and subsequent activation of

its kinase domain. Activation involves association of tyrosine phosphorylated

residues on RET with PLC-g1, Shc, Grb2, and others through to Ras-ERK, PI3K.

The reader will recognize that the pathways involved resemble those activated

by neurotrophins. Heparin sulphate proteoglycans concentrate extracellular

gradients of GDNF to facilitate its interaction with GFRa and RET [597]. Signaling

without RET activation may also occur [58]. There are also interesting and

important interactions of the neural cell adhesion molecule (NCAM) with GDNFs

and its GFRa receptors [597]. Similarly, GDNF can act synergistically with laminin

to promote outgrowth of IB-4 NGF unresponsive, TrkA negative adult sensory

neurons [698]. Like NGF, GDNF also activates RAC GTPase in growth cones to

facilitate lamellipodia formation [202].

GDNF is expressed at only low levels within intact peripheral nerve or ganglia

but it is upregulated in Schwann cells after injury. Both GFRa-1 and GDNF levels

rise at sites distal to a nerve trunk injury, illustrating another form of axon–SC

interchange to support regeneration. GDNF, elaborated by SCs, signals axon

outgrowth through the upregulated expression of its receptor on local axons.

In contrast, mRNAs for neurturin, artemin, and persephin can be identified in

peripheral nerve trunks and ganglia but do not seem to be altered by injury.

Similarly, the receptors GFRa2 and RET mRNAs are also present in ganglia and

peripheral nerve but are not altered by injury for up to 2 weeks. There was a late

rise in GFRa2 in chronically denervated nerve stumps at 3 and 6 months after

transection. GFRa3 was upregulated in proximal stumps but not in distal stumps

or ganglia after injury [277,278].

When expressed, GDNF acts as a trophic factor for motor neurons, highlighted

by a recent review by Bohn [50] entitled “Motor neurons crave glial cell line-

derived neurotrophic factor”! A number of studies have identified rescue of the

retrograde loss of motor neurons after axotomy or rescue of motor neurons in

experimental ALS by GDNF. For example, in neonatal mice that overexpress GDNF

in astrocytes, facial motor neurons were protected from retrograde loss [777].

Other neurons also “crave” GDNF to a greater or lesser degree. For example, GDNF,

neurturin, and artemin support survival and differentiation of sympathetic,


parasympathetic, sensory, and enteric neurons [597]. One interesting feature

of artemin is its expression along blood vessels where a chemotactic gradient

of the molecule attracts innervation by sympathetic perivascular axons [372].

This may allow formation of sympathetic neuroeffector junctions on blood

vessels. Artemin can also increase neurite outgrowth in sensory neurons. Its

receptors (GFRa3) are expressed in about a third of adult rat DRG neurons,

preferentially in small peptidergic and IB-4 TrkA negative neurons. Artemin also

protects sensory neurons in vivo. Administered intrathecally after axotomy, arte-

min prevented changes in peptide expression, upregulation of injury markers

(ATF-3), and declines in unmyelinated axon conduction velocity proximal to the

injury site [40].

Fibroblast growth factors (FGFs) were originally identified through their ability

to stimulate fibroblast proliferation. Two main subtypes, FGF-1 (acidic FGF) and

FGF-2 (basic FGF) had been recognized but recently the number of FGFs has

increased dramatically to 22 [174]. In the extracellular matrix FGFs may be

associated with extracellular heparin or heparan sulphate proteoglycan (HSPG).

As ligands, FGFs operate on four main tyrosine kinase receptors known

as FGFR1–4. FGFR-1 is expressed on neurons and FGFR-2 and 3 are expressed

by glial cells. Two downstream docking proteins FRS2a and FRS2b stimulate

the Ras/MAPK and PI3K-Akt signaling pathways indicating a convergence with

actions of other growth factors discussed above. Eight members of the FGF

family, FGF 2, 5, 7, 8, 9, 10, 13, and 14, have been identified in rat lumbar 4

and 5 dorsal root ganglia (DRGs) [396]. Specifically, FGF 2 and 8 are expressed in

sensory neurons [309,396,545,558,595,666] and FGFRs are upregulated in DRG

neurons following injury [327]. FGF-1 is a known survival factor for injured rat

spinal neurons [384] whereas FGF-2 mRNA is increased following embryonic, but

not adult spinal cord injuries [563]. FGF-2 and FGFR1–3 mRNAs were all increased

in crushed sciatic nerves [238].

Transforming growth factors b (TGFb) are critical trophic molecules that comprise

over 50members. TGFb1 and 2 are expressed in the peripheral nervous system and

in SCs. While TGFs interact with other growth factors, their actions differ since

TGFbs inhibit proliferation. TGFb2 and its receptors (Type I and II) are anterogra-

dely and retrogradely transported inmotor neurons [310]. A combination of TGFb1

and forskolin, a molecule that raises cAMP levels, increased the growth of motor

neurons regenerating into chronically denervated nerve stumps [657]. TGFb1

is upregulated in the distal stump of injured peripheral nerves by SCs [587]. Bone

morphogenetic proteins (BMPs) are a subfamily of TGFbs that support bone growth

but are also expressed in peripheral neurons and SCs. BMP-2 is specifically

expressed in motor neurons. BMPs engage serine/threonine kinase receptors

(BMPRII, ALK2,3 and 6) and activate Smad proteins. BMP-4 and BMP-6 potentiate


NT-3 and neurturin-mediated neurite outgrowth and survival in chicken

embryonic sympathetic and nodose ganglia [11]. Local administration of BMP-2

improved facial nerve axon regeneration 4 weeks following crush in rabbits [724].

Epidermal growth factor (EGF) plays critical roles in early behavior of neural

precursor cells but may have a lesser role in adult nerve regeneration [372].

Heregulin polypeptides, also known as neuregulins were previously introduced

in Chapter 5, and share an EGF domain. The NRG gene splice products have

included previously identified proteins called neu differentiation factor (NDF),

heregulin, acetylcholine receptor inducing activity (ARIA), and glial growth

factor (GGF) [272]. Neuregulins are largely growth factors for SCs. Platelet-derived

growth factor (PDGF) and its receptors are synthesized by neurons and SCs [372].

Erythropoietin (EPO) is released by peripheral nerve SCs in response to nitric

oxide and signals EPO-R receptors on peripheral axons. The receptors are also

expressed on sensory neurons in DRGs [87]. EPO protects axons from degener-

ation, prevents sensory neuron apoptosis, and promotes axonal regeneration

[87,88,338].

Insulin, and closely related insulin-like growth factors (IGFI and II), support the

metabolism, growth, differentiation, and survival of neurons and other cells.

Insulin receptors (IRs), consisting of an a and a b subunit, are widely expressed

on peripheral sensory and motor neurons. Ligand binding of the a subunit leads

to tyrosine autophosphorylation of the b subunit, activation of the catalytic

domain, and phosphorylation of substrates, including the insulin receptor sub-

strate (IRS; IRS-1 and IRS-2) proteins and Shc [546,738]. Phosphorylation at mul-

tiple IRS serine/threonine and tyrosine sites generates a signaling scaffold or

docking pathway [737]. As might be expected, IRS-1 is frequently colocalized with

insulin and IGF-1 receptors, and therefore activates a variety of familiar messen-

gers including: PI3K-Akt, Shc,Grb2,S6 kinase, PKCe kinase, MAP2 kinase, Raf1

kinase, and cfos [177,186,257]. PI3K-Akt signaling, activated by insulin and other

growth factor receptor tyrosine kinases, blocks apoptosis through interactions

with BAD, caspase-9, NFkB, and the forkhead transcription factor FKHRL1

[70,163,496] (for reviews see [189,251]). At low nanomolar physiological concen-

trations, insulin stimulates neurite outgrowth using insulin receptors, while at

supraphysiological doses, insulin also binds to Type 1 IGF-Rs.

IGF-1 receptors (IGF-R) are present on neurons and help to retrogradely

transport IGF [125]. Expression of IGF-1 and IGF-R occurs in small (<25mm)

diameter DRG neurons [125]. IGF-1Rs are also expressed on Schwann cells [109]

and promote myelination [110]. Like insulin, IGFs bind to a subunits of IGF-Rs

extracellularly to activate the intracellular b subunit, which induces a sustained

tyrosine phosphorylation of Shc associated with Grb2 while simultaneously

activating ERK [341]. IGF-1 also induces a transient tyrosine phosphorylation of


IRS-2 and an association of IRS-2 with Grb2. IRS-2, and Grb2, but not Shc are

concentrated at the tip of the growth cone where the IRS-2–Grb2–PI3K complex

regulates extension andmembrane ruffling [341]. Like insulin, IGF-1 also protects

neurons against oxidant stress and apoptosis [589]. Thus, IGF-1 also upregulates

uncoupling protein 3 (UCP3), a member of the mitochondrial transporter super-

family found within the inner membrane of mitochondria [245]. UCP3, in turn,

regulates the production of reactive oxygen species (ROS) in mitochondria to

protect neurons from oxidant damage.

Pleiotrophin (PTN) is a member of the mid-kine (MK) family of heparin binding

growth factors [313]. PTN is also known as heparin binding growth-associated

molecule or heparin binding neurotrophic factor. Both PTN and mid-kine, the

other family member, are cysteine- and basic amino acid-rich proteins with 50%

amino acid homology. They operate on several receptors including low density

lipoprotein receptor-related protein (LRP-1, see Chapter 5), anaplastic lymphoma

kinase (ALK), protein tyrosine phosphatase (PTP), and N-syndecan. Both PTN and

MK promote neurite outgrowth [318,402]. In the peripheral nervous system, PTN

can be identified in nerve stumps distal to crush in mice and is localized to SCs,

macrophages, and endothelial cells [49]. PTN is upregulated in acutely but not

chronically denervated distal sciatic nerve stumps [467].

Hepatocyte growth factor (HGF), also known as “scatter factor” because it can

induce epithelial proliferation, is a pleiotropic molecule that acts through the

Met tyrosine kinase receptor (see review [435]). It has been most extensively

examined during development. HGFs offer their support to spinal motor

neurons through self-synthesis and an autocrine loop [372,435]. HGF is also

expressed in adult denervated Schwann cell populations [280], where it acts

as a mitogen [361]. HGF may have synergistic interactions with NGF by

acting through shared downstream pathways such as PI3K, PLC-g, and Ras/MAPK

pathways, but it may not enhance the actions of other neurotrophins [435].

Macrophage stimulating protein (MSP) is a growth factor closely related to

HGF [372].

Vascular endothelial growth factor (VEGF) [93,649] is widely known for its role in

vascular angiogenesis, but it also has direct actions on peripheral neurons. VEGF

is expressed in a subset of small ganglia neurons and its receptor fetal liver

kinase receptor (flk-1) is found in sensory ganglia neurons, growth cones as well

as SCs. VEGF stimulates axon outgrowth from adult explant cultures of sensory

ganglia through the MAPK pathway, and also increases the survival of neurons

and satellite cells [634]. VEGF also promotes SC proliferation; two studies have

examined in vivo outgrowth of axons from nerve transection sites using local

exogenous VEGF treatment. In one, SC but not axon outgrowth increased [635],

whereas in the second study, both were enhanced [274]. While speculative, it is


an attractive idea that VEGF could operate to coax concurrent outgrowth of

axons and blood vessels (since it signals both) during some nerve injuries.

How neurotrophins and other growth factors impact regeneration

Neurotrophins ligate receptors both in the periphery and growth cones,

and at the level of the cell body [679]. Within ganglia, they are released into the

extracellular milieu by neighboring neurons or satellite cells and interact with

perikaryal receptors. In contrast, retrogradely transported neurotrophins from

the periphery in endocytotic vesicles may offer direct signals to the nucleus of

the neuron [262,727]. Some peripheral signaling, however, is also completely

local, without the necessity for transport to the nucleus (Figure 9.6).

Ignoring local signaling for the moment, neurotrophins and other growth

factors influence the repertoire of nuclear genes expressed in response to axot-

omy, also known as RAGs (regeneration associated genes; see Chapter 5). These

are complex interactions, however, since some trophic factors seem to shut off

RAGs (e.g., NGF applied to axotomized sensory axons) [707,708] suggesting that

they compensate for factors no longer available from axoplasmic transport.

In contrast, other neurotrophins applied, for example, in the CNS, contrarily

Autocrine

Target tissue

Paracrine

Schwann cell

Retrograde transport

Figure 9.6 Illustration of sources of growth factor support. (Reproduced with

permission from [782].) See color plate section.


increase RAG expression (e.g., BDNF applied to sectioned CNS rubrospinal

neurons) [216,350]. This paradox may simply speak to the differences between

central and peripheral nerve regeneration. In the PNS, RAGs show no hesitancy

in being upregulated after an axonal injury. In contrast, CNS axotomized

neurons appear to experience dramatic delays recognizing that their axons are

damaged and require a response! The delayed RAG response by CNS neurons is

one of several differences between central and peripheral regeneration

(see [179]).

In peripheral sensory neurons NGF and GDNF (see below) delivered intrathe-

cally to access perikarya, each prevented rises in the RAG known as ATF3, a stress-

activated transcription factor [28]. As one might expect, the particular neurons

impacted by these growth factors were selective. Thus, NGF prevented ATF3

upregulation in small- to medium-sized TrkA neurons, whereas GDNF impacted

ATF3 in neurons known to be GDNF responsive and that express the receptor

molecule P2X3 (also known as IB4 GDNF responsive neurons). Whether down-

regulating the ATF3 RAG is beneficial for regenerating neurons is unclear. ATF3

is induced by JNK and forms heterodimers with leucine zipper transcription

factors. Its actions have both protective and detrimental properties. For example,

overexpressing ATF3 in adult sensory neurons increased neurite outgrowth

in the presence of both NGF and GDNF [612]. The case of ATF3 is significant in

that it illustrates how the RAG program can be substantially altered by growth

molecules.

At the growth cone, with minutes of its application, NGF can signal changes in

actin dynamics, specifically its localization to leading tips and its polymerization

[226]. Through this mechanism, only small amounts of NGF are required to

initiate changes in lamellipodia behavior [669]. Moreover, the rapid response to

NGF can occur prior to the neurotrophin even reaching the central zone of the

growth cone let alone the nucleus! Studies using Cy3-labeled NGF have illus-

trated that NGF signals begin on ligation of its receptor before internalization.

Other factors and second messengers influence how neurotrophins signal

locally at the growth cone or at the nucleus after retrograde transport. NGF

may activate Erk1, Erk2, and Erk5 in the cell body when it is applied to distal

axons, whereas only Erk5 is activated if NGF is applied to the cell body [726].

Endosomal recycling of TrkA is important for both retrograde transport

of endocytotic vesicles containing neurotrophins and for local NGF signaling.

Several molecules may influence Trk endosomal recycling. One example is Rab7,

a member of the RabGTPase family that controls both endosomal trafficking

but also regenerative signaling of TrkA. Mutations of Rab7 are associated with

inherited polyneuropathy [599]. Another molecule that appears important for

retrograde degradation-resistant TrkA and TrkB transport by endosomes is


known as Pincher, the pinocytotic chaperone. Pincher mediates internalization

of activated Trks within endosomes from plasma membrane ruffles. These are

later converted to multivesicular bodies that can signal continuously during

retrograde transport retrogradely to the nucleus, where they then activate

Erk5 to promote neuron survival [701].

Neurons cultured in Campenot chambers (see Chapter 4) allow segregated

testing of signaling in perikarya or axons that are bathed in separate media [90].

Concentrations of NGF placed into axon side chambers but not in the perikaryal

media, directed axons to grow toward the NGF source [91]. NGF had synergistic

actions with NT-3 but not with BDNF in promoting axon outgrowth into Campenot

side chambers.

Local protein synthesis plays a critical role in influencing axon behavior

during regeneration (see Chapter 5). While not fully explored, neurotrophins

and other growth factors likely have a substantial influence on axon protein

synthesis. For example, NGF and BDNF increase b-actin, peripherin, and vimentin

mRNA transport into axons [742]. Beyond their actions on transport, however, it

is likely that growth factors can also promote translation. The subsequent

localization of newly synthesized proteins to particular parts of the growth

cone may also be influenced by growth factors. This localization, such as the

asymmetric distributions of microtubular or actin-associated proteins helps to

determine the direction of growth. Thus neurotrophins also have a significant

bearing on guidance, or tropism and chemotaxis by influencing growth cone

dynamics [547,699]. NGF, NT-3, NT-4/5, and BDNF can generate chemoattractive

growth cone turning, and in some instances, chemorepulsion.

Overlapping expression of growth factors and their receptors after injury

Neurotrophins do not influence axons in isolation. Their behavior is best

understood by examining the cells and timetables that generate them. Some

neurotrophins are expressed in neurons, particularly BDNF, where they act as

autocrine trophic factors [579]. After peripheral nerve injury, however, Schwann

cells and macrophages are thought to be the primary sources of growth factors,

though other cells also contribute [266]. For example, mast cells express NGF and

BDNF in injured nerves [787]. The likelihood that neighboring cells will release

NGF or other neurotrophins, in turn, depends on other stimuli such as IL-1, other

cytokines, TGF-b, Wnt, and some specific hormones [76,408,579].

Neurotrophin release and receptor expression after injury are also complex

(see review [58]). In axotomized motor neurons, BDNF and IL-6 mRNAs rise, peak,

and fall within the first 7–10 days after injury. After the peak of expression, BDNF

returns to baseline at a slower rate than IL-6. There are also differences in IL-6

expression between facial and femoral motor neurons. In contrast, NT-3 and


NT 4/5 mRNAs decrease in motor neurons after injury. These variations in

expression are not easily understood.

It is essential that growth factor receptors are available to accommodate

higher loads of ligand after injuries. TrkB, RET, GFR-a1, and LIFR have sustained

rises within motor neurons beginning with days of axotomy. Rises in both ligand

and receptor are also essential to support autocrine self-signaling.

Axons must regrow into distal nerve stumps to reach target organs and effect

functional reinnervation. The elaboration of growth factors by non-neuronal

cells in the denervated distal stump can maintain axon ingrowth before their

contact with target organs. After the first week following denervation, slow rises

were detected for mRNAs of BDNF, NGF, and NT-4/5. The largest rise was in BDNF

with low level sustained rises in p75, and truncated TrkC, TrkB [58,203]. SCs also

expressed truncated TrkB receptors that were downregulated in the distal stump

after denervation. Both NGF and BDNF were detected in SCs 1–12 days following

axotomy in the distal stump following a facial nerve trunk injury [651]. FGF-2 and

IGF-1 were detected in distal stumps at days 1–8 in two separate peaks at days

2 and 6. In situ hybridization studies confirmed that mRNAs of NGF, BDNF, and

FGF-2 were present in SCs and probably also macrophages [238,651]. GDNF and

GFRa-1 rose within distal stumps in the first week, then had slow declines. In

separate work, mRNAs GFRa-1 continued to rise later, peaking 3 months after

injury. Similar, but less robust late rises in GFRa-2 were also identified. There

were no rises in neurturin, artemin, or persephin or in the receptors GFRa-3 or

RET in distal nerve stumps after transection [278].

In the first week after nerve trunk injury, differing timetables in the

responses of other neuropoeitic growth factors in non-neuronal cells of the distal

stump were also evident. There were rapid short lived rises over the first few days

of IL-6 compared to a sustained rise in LIF and rises in IL-6R. Alternatively, CNTF

declined and disappeared, then increased after 10 days. A small rise in the first

few days was noted for the neuropoeitic growth factor receptor subunit gp130.

Taken together, these features illustrate a complex and overlapping set of

changes operating through several mechanisms. They also indicate substantial

redundancy in the support of regenerating axons. There are major roles for early

neuronal and perhaps SC elaboration of growth factors and their receptors, and

somewhat later release by SCs, macrophages, and mast cells in the distal stump.

Not all SCs behave alike in a distal denervated nerve stump. Ventral roots, for

example, upregulated pleiotrophin and GDNF but not NGF, BDNF, VEGF, HGF, or

IGF-1. These latter factors, in contrast, were upregulated in denervated cutane-

ous nerves [280]. Target organs also generate growth factors that guide axons

to them. For example, denervated muscle supplies NGF, NT-3, and NT-4/5[233].

Finally, following transection nerve injuries, axons grow out from the proximal


nerve stump before they can reach distal stumps containing growth factors.

While very early axon sprouts likely do not require trophic support, sustained

outgrowth does. Their support may come from proximal stumps where macro-

phage influx, SC plasticity, and mast cell proliferation occur. Each may contrib-

ute growth factors to support outgrowing axons [106,807]. Both NGF and BDNF

were detected in SCs of proximal stumps of transected nerves at 3–6 days

following axotomy. bFGF and IGF-1 were detected in the proximal stumps at

days 1–8 in two separate peaks at days 2 and 6. GDNF was detectable in the

proximal nerve stump between days 1 and 6.

Clearance of myelin-related and other regeneration inhibitors from the

distal nerve stump is thought to be rapid. Despite these efficiencies, however,

it is possible that some inhibitors persist in this milieu. These may include

chondroitin sulphate proteoglycans (CSPGs), semaphorins, myelin associated

glycoprotein (MAG), and others. Growth factors can overcome such inhibitors.

Priming sensory neurons with BDNF or GDNF in vitro blocked inhibition of axon

regeneration by MAG. Their actions involved increases in cAMP and PKA [84].

In xenopus motor neurons, BDNF, GDNF, NT-3, NT-4, and interventions that

increased cAMP in neurons [548] enhanced axon outgrowth. Paradoxically these

beneficial interventions also inhibited the formation of neuromuscular junc-

tions. Thus while growth factors support initial outgrowth of motor axons they

may not necessarily facilitate proper maturation of new connections to muscles.

Some specific studies of growth factors and their impact on nerve regeneration

Investigators have not discounted the possibility that growth factors

might offer exciting avenues for the treatment of nerve injuries. Growth factors,

combinations of growth factors, or growth factor mimics have been delivered in

a variety of ways to peripheral neurons. Some studies have emphasized exogen-

ous delivery, while others have explored endogenous expression. This section

will briefly examine some of those studies. A review of how specific regeneration

conduits deliver growth factors to promote regeneration has been published

[555]. Briefly, these include: (i) providing soluble growth factors in the lumen of

conduits connecting transected nerves; (ii) growth factor delivery from a matrix

inside the lumen of a conduit; (iii) growth factor delivery from osmotic mini-

pumps or other approaches; (iv) delivery from microspheres; (v) delivery from

the conduit wall. A fibronectin “mat” that is wrapped around the two stumps of

the nerve has also been used. When impregnated with NGF, it enhanced regener-

ation in primates [4]. Gradients of growth factors may be important during early

regeneration. These may be set up through the design of embedded conduits or

by the delivery approach [331]. As already discussed, SCs or stem cells have been

implanted between stumps of transected nerves. Pfister et al. [555] discuss the


pros and cons of various approaches. For example, conduits filled with material

may exert a physical barrier to regrowth. Gap lengths of greater than 10mm

offer particular challenges to regrowth. Finally optimal conduit design should

include biodegradability to limit later constriction and an ongoing foreign body

inflammatory reaction [35,36,326].

Some of the most rigorous and intriguing investigations addressing endogen-

ous NGF were carried out over 15 years ago by the Diamond laboratory [139,141].

The investigations addressed how cutaneous nerves, isolated on the dorsal trunk

of the rat, undergo collateral sprouting of nociceptive axons when adjacent

territories are denervated (see Chapter 4). The approach focused on how axons

that might be NGF responsive (cutaneous nociceptive axons) might collaterally

sprout and enlarge their territories. Subcutaneous injections of antiserum to

NGF prevented the expansion of their cutaneous fields. By contrast, Aa axons

subserving light touch did not sprout, consistent with their insensitivity to NGF.

When collateral sprouting was already in progress, later application of anti-NGF

halted further growth, whereas stopping anti-NGF injections allowed sprouting

to resume. Normally, there was an estimated latency of the growth of these

receptive fields of 8–10 days, thought to be due to a delayed rise in NGF content

in the denervated skin target tissue. In additional work, intradermal NGF injec-

tions evoked collateral sprouting in normal fully innervated skin, increased the

rate of sprouting into denervated skin, and restored sprouting arrested by anti-

NGF. A final critical finding was that collaterally sprouting axons in the skin

followed residual perineurial tubes left over from the denervated nerve trunks.

Diamond and colleagues also noted that regenerative axon sprouting,

in contrast, completely ignored anti-NGF injections. The experiments used the

same isolated dorsal cutaneous nerves, but the remaining T13 branch was

crushed and subsequent reinnervation was assessed. Recovery (the rate and area

of reinnervation) was independent of anti-NGF. Other types of injuries, such as

transection were similarly insensitive to NGF. In some rats given anti-NGF after

two separate injuries, collateral sprouting could be halted in one territory,

whereas regenerative sprouting was unchanged in another. Neither Aa, Ad, nor

C fibers had regeneration blocked by anti-NGF nor accelerated by exogenous

NGF. In combination, this series of experiments indicated that endogenous NGF,

synthesized by the skin, supported collateral reinnervation of nociceptive axons

but had no bearing on regenerative sprouting.

The impact of exogenous NGF on regeneration does remain controversial.

Several investigations have noted somewhat surprising actions of NGF alone,

or together with other growth factors, on the regeneration of broad categories

of peripheral nerve axons. For example, NGF cDNA delivered intrathecally with

a polyethylenimine complex, improved regeneration (density and caliber of


myelinated axons) of sciatic nerves across conduits bridging transections [722].

NGF immobilized by heparin in a fibrin matrix enhanced the regrowth of axons

through a conduit connecting a 13mm sciatic nerve transection gap [382]. NGF

and GDNF enhanced the proportion of process bearing adult sensory neurons

and total neurite length in vitro [212]. BDNF and NT-3 had inhibitory effects.

It seems difficult to understand how NGF might support neurons that do not

express TrkA. As discussed above, NGF-p75 signaling may substitute, or it may be

that in some experiments the NGF preparation contained the proneurotrophin

precursor. Another possibility is that NGF directly signals SCs to migrate out

from injury sites first, which in turn seduce axons to follow them [19].

The idea that BDNF influences regeneration is less controversial. BDNF rescues

adult motor neurons following injuries [347,509,510]. Similarly BDNF is capable

of reversing axotomy-related declines in the motor neuron molecules as ChAT

and AchE (see Chapter 5) [347]. The actions of BDNF during nerve regeneration,

however, are variable. Exogenous BDNF did not improve some indices of func-

tional recovery after transection and repair [476,621]. BDNF did promote axon

regeneration into a chronically denervated distal stump but higher doses had an

opposite action and reduced reinnervation [56]. Higher doses of BDNF might

antagonize BDNF–TrkB signaling through stimulation of p75. Antibodies to BDNF

reduce the number and caliber of regenerating myelinated axons and the dis-

tance of their regrowth (pinch test) in crushed sciatic nerves [772]. Similarly,

mice heterozygous for TrkB (þ/–) (homozygous deletions in mice are neonatal

lethal) have normal numbers of intact but attenuated numbers of regenerating

motor neurons [55].

Endogenous BDNF may be released by electrical stimulation of nerve

trunks (see Chapter 10), an intervention discovered to enhance axon regener-

ation and preferential motor reinnervation [7]. Thus, TrkB þ/– heterozygous

knockout mice did not benefit from stimulation. The impact of electrical stimu-

lation through BDNF/TrkB signaling also includes a change in SC phenotype.

Stimulation upregulates the HNK-1 carbohydrate epitope of premyelinating

SCs that associate with motor axons and may support preferential motor rein-

nervation [162].

CNTF and BDNF supplied together and linked to collagen tubes had only

minor benefits on axon maturation or functional recovery after transection

and repair [271]. A chimeric pan-neurotrophin made by combining the active

domains of NGF, BDNF, and NT-3 and linked to the BDNF promoter increased

axonal sprouting and regeneration of myelinated axons and promoted motor

axon regeneration and sensory reinnervation of the skin [294]. Focal application

of NGF, BDNF, and IGF-1 reduced inappropriate collateral sprouting of transected

motor neurons into incorrect branches [651].


NT-3 applied by fibronectin mat at a site of nerve repair increased axon

outgrowth across a 10mm gap and reduced denervation muscle fiber atrophy

[628,640,641]. Several studies have identified an impact of NT-4/5 on regeneration

of rat sciatic nerves [170,627,765]. NT-4/5, applied by fibrin glue to a repaired

nerve transection site improved walking tracks, myelinated axon diameter,

myelin thickness, and numbers of axons up to 60 days 10mm beyond the injury

site [765]. The outgrowth of YFP fluorescently labeled peroneal axons entering

grafts from NT4/5 –/– or NT4/5 –/þ mice was of interest [170]. Both transgenic

grafts were less supportive of outgrowth than wild-type grafts. BDNF þ/– grafts,

however, had no impact on regrowth. In grafts treated with fibrin glue containing

BDNF or NT4/5, the regenerative deficit in the NT4/5 –/– grafts was reversed and

regeneration exceeded that of wild-type grafts. Finally, in p75 (–/–) homozygous

mice, with a reduced number of intact sensory neurons during development,

there were greater numbers of regenerating motor axons [55] but no overall

impact on other facets of peripheral nerve trunk regeneration (Karchewski,

Verge, Zochodne, unpublished data).

GDNF promoted reinnervation of chronically denervated nerve trunks but

had little impact on early regrowth after transection [57]. Mice lacking IL-6 had

delayed regeneration of sensory axons following crush [298,778]. Similar but

more persistent delays were identified in mice lacking CNTF [760]. Exogenous

CNTF potentiated peripheral nerve regeneration by increasing the distance of

axon outgrowth and maturation of myelinated axons during regeneration [591].

The roles of FGFs during peripheral nerve regeneration are diverse (see review

[237]) and their impact depends on the receptor expressed. As suggested above,

FGF-2 and FGFR1–3 were increased proximal and distal to crush sites within

peripheral nerves and FGF-2 and FGFR3 were increased in DRGs following axot-

omy. Thus, the upregulation of FGF-2 growth factor after injury is well positioned

to support regeneration. It promotes neuronal survival and neurite outgrowth

when given exogenously or in vitro. Alternatively, FGF-2 binding to FGFR3 is also

associated with neuron apoptosis and FGF-2 and FGFR3 knockout mice are

protected from axotomy-induced apoptosis. FGF-2 acting on FGFR1/2 may inhibit

myelination and SC proliferation. The FGF field therefore promises to identify

potent and direct but overlapping actions, sometimes conflicting, on neurons

and their supporting cells.

Insulin and IGF-1 stimulate survival, neurite outgrowth, and other responses

in adult sensory neurons [577]. Unlike other growth factors, insulin and IGFs

circulate in the bloodstream and may exert direct actions without the prerequis-

ites of target expression or transport. Insulins have potent actions in vivo

[75,323,653,754]. Insulin accelerates nerve regeneration, especially when applied

intrathecally to access perikarya [689,754] and reverses retrograde atrophy and


axon loss after nerve injury. Insulin directed to sensory neurons intrathecally

also promotes reinnervation of the skin in experimental diabetes [690]. IGF-II

expressed in muscle stimulates sprouting of motor axons during partial dener-

vation or botulinum paralysis [219,220].

PTN increased motor axon outgrowth from spinal cord explants and in doing

so formed “miniventral rootlets.” It also protected neonatal motor neurons

from axotomy-induced dropout [467]. PTN similarly increased axon regeneration

across a transection injury and accelerated the reconnection of motor axons to

endplates in rats when delivered locally by HEK cells. All of these properties are

believed to be mediated by ligation of the ALK receptor.

The control of remyelination by regenerating axons is also directly influenced

by specific neurotrophins. While not extensively studied during adult nerve

regeneration in vivo, such actions have been documented during development

or in vitro. Transgenic mice with BDNF overexpression have persistent (into

adulthood) increases in myelin content and thickness [683]. NGF promotes mye-

lination by SCs [96]. TGFb1 may regulate the stability of mature myelin [130].

Mice with a null mutation of the TGFb1 gene have impaired myelin development

rendering structurally abnormal myelin sheaths [130]. A combination of PDGF

and IGF-1 enhances myelination but does not improve regeneration across an

injured spinal cord [539].

A number of other endogenous or exogenous trophic-like agents have mim-

icked actions of growth factors. For example, “neotrofin” is a synthesized purine

derivative that mimics some of the actions of NGF. Neotrofin induced cutaneous

collateral nerve sprouting [283]. Other examples include nucleosides, nucleot-

ides, arachidonic acid, and others [328,502]. Through a screening process, small

nonpeptide molecules that are p75 ligands have been discovered that support

neuron survival without an interaction with Trks [444].

Summary

A number of protein families influence the regeneration of peripheral

neurons. The most widely studied are the neurotrophins that operate through

specific tyrosine kinase receptors. Others include the neuropoeitic cytokines,

GDNFs, insulin-likemolecules, and a series of growth factors better known for their

actions on other cell types. Many growth factors that impact peripheral neurons

activate overlapping signaling cascades that eventually involve PI3K-Akt, Ras/ERK,

and PLC-g1. Through knockout studies illustrating endogenous actions or investi-

gations using exogenous application, many have potent actions during peripheral

nerve regeneration. Their supportive actions may not be confined to neurons

but may also operate through the support of SCs and other supporting cells.


Suggested reading

Boyd, J. G. & Gordon, T. (2003). Neurotrophic factors and their receptors in axonal

regeneration and functional recovery after peripheral nerve injury. Molecular

Neurobiology, 27 (3), 277–324 [58].

Landreth, G. E. (2006). Growth factors. In Basic Neurochemistry, ed. G. J. Siegel,

R. W. Albers, S. T. Brady, & D. L. Price. New York: Academic Press, pp. 471–484 [372].

Levi-Montalcini, R. & Hamburger, V. (1953). A diffusible agent of mouse sarcoma,

producing hyperplasia of sympathetic ganglia and hyperneurotization of

viscera in the chick embro. Journal of Experimental Zoology, 123, 233–287 [390].

205Summary

10

The nerve microenvironment

Nerve regrowth has a context, a microenvironment replete with mole-

cular, physical, and other determinants of success. Some are roadblocks, others

are partners. The growth factors considered in Chapter 9 are only part of

the story of how nerves fare during regeneration. In this chapter, we consider

a number of other salient features of the regenerative milieu that influence

regeneration. These range from adhesion and basement membrane molecules, to

novel signaling systems to interactions within altered environments as occur

during aging, diabetes mellitus, or entry into the spinal cord.

Adhesion molecules

Cell adhesion molecules (CAMs) represent the interface between axons,

SCs, and the basement membrane. They provide the essential extracellular

“clutch” domains that allow differential movement of membranes and mol-

ecules along one another. During nerve regeneration, adhesion molecules

permit axon growth with basement membranes, along guiding SC processes, or

with other fasciculating axons guided by pioneer axons. In turn, the linkages of

CAMs to the intracellular cytoskeleton influence cell mobility. To accomplish

their task, CAMs can interact with each other (homophilic) or other molecules

(heterophilic). These interactions vary in adhesive strength depending on the

location and intensity of their expression, distribution on a given cell, and

electrical charge.

Three major families of CAMs are recognized. The immunoglobulin (IgCAMs)

family contains extracellular b sheets that resemble the variable or constant

domains of immunoglobulins (Igs). This family is further subdivided into three

types, depending on the type of Ig domains and the presence of fibronectin-like

206

or other extracellular domains. Some have intracellular signaling kinase

moieties or glycosylphosphatidylinositol (GPI) anchors. They can bind in a homo-

philic or heterophilic fashion. Overall examples include the myelin proteins

P0 and myelin associated glycoprotein (MAG), Thy-1, neural cell adhesion mol-

ecule (NCAM), and L1. Integrins, considered below, are divalent cation-dependent

(Ca2þ or Mg2þ) adhesion molecules that bind in heterophilic fashion. Cadherins

are calcium-dependent adhesion molecules that usually bind in a homophilic

fashion, and have an amino terminal extracellular domain, a single transmem-

brane domain, and a carboxyl-terminal intracellular domain. Several different

types exist, named for the first tissue in which they were identified (though now

known to be more widely expressed): N-cadherin or neural cadherin, E-cadherin

or epithelial cadherin, and P-cadherin or placental cadherin. Cadherins generate

adhesion through extracellular lattice “plates” that form an adhesive dimer

across facing cell membranes (see review [121]).

Overall, there are several adhesion molecules expressed by neurons, SCs, and

the adjacent basement membrane (BM) that influence axon outgrowth [440].

NCAM is upregulated by SCs distal to an injury site [439]. NCAM is also, however,

expressed on axons. Axonal plasticity, in turn, can be enhanced by the addition

of a polysialic acid (PSA) post-translational modification. For example, regener-

ating motor axons destined to motor pathways express high levels of NCAM–PSA

[190]. NCAM may facilitate neurite outgrowth through an interaction with the

FGF receptor [439,718].

The L1 adhesion molecule is expressed by neurons and SCs and is upregulated

by NGF [144,316,613]. Like NCAM, L1 is also upregulated in SCs of the distal

stump after nerve injury. HNK-1 (L2) is expressed in SCs and BM and selectively

facilitates motor axon outgrowth [441]. To accomplish this task, HNK-1 is only

present in SCs associated with motor fascicles. In contrast, sensory fascicles

express NCAM in SCs associated with unmyelinated sensory axons [592].

Overall, it is apparent that several redundant types of adhesion molecules

foster interactions between axons and SCs. Adhesion molecules are likewise

important for later maturation events of peripheral nerves including fasci-

culation and myelination. For example, P0 and MAG have particular roles in

myelination.

Basement membranes

Extracellular basement membranes (BMs) provide the pavement or

handholds for axons to move forward. BMs, in turn, require appropriately

expressed axon molecules, such as integrins to accurately support and guide

axons and growth cones. Hence a two-way cooperation between “fixed” ligands

207Basement membranes

and their receptors is a fundamental requisite. The menu of BM constituents that

new SCs and axons are exposed to varies with the type of injury, its duration, and

the extent of regrowth. SCs elaborate most BM molecules, but their own migra-

tion and guidance may depend on pre-existing molecules. Below we discuss the

major extracellular BM constituents identified within peripheral nerves.

Laminins are heterotrimeric glycoproteins with one a, one b, and one g chain.

Up to 14 isoforms from five a, three b, and three g subunit types have been

described. They are cruciform, or cross-like in structure. Endoneurial basal

laminae of peripheral nerves contain laminin-2 (merosin; a2b1g1) and laminin-8

(a4b1g1) (Figure 10.1). Both appear critical to aspects of nerve regeneration, but

they interact with different receptors. Laminin is a preferred axon outgrowth

substrate. This was demonstrated by Gundersen during development [243].

In an elegant series of experiments, neonatal chick sensory neurons were given

the opportunity to sample and grow along several choices of BM laid down in

strips. Axon outgrowth clearly favored laminin strips rather than fibronectin

or collagen IV.

Since laminin is essential for peripheral nerve development, severe abnormali-

ties in nerve structure occur when there is abnormal expression of laminin.

For example, mice lacking a4 of laminin-8 have abnormalmyelination with higher

G ratios indicating myelin thinning. They also have abnormal radial sorting

of axons within SC basement membranes [716]. Despite their abnormal develop-

ment, however, mice lacking a4 of laminin-8 had normal axon regeneration

after crush.

Figure 10.1 Immunohistochemical image of the relationship between laminin

(labeled with an anti-laminin antibody, green) and axons (labeled with an

anti-neurofilament antibody, red) in a peripheral nerve stump. Laminin is associated

with basement membranes of SCs surrounding axons and with blood vessels.

[Bar ¼ 50 microns] (Image taken by Chu Cheng, Zochodne laboratory.)


208 The nerve microenvironment

Although SCs have longer processes when grown on laminin-8, embryonic

motor neurons extend on laminin-2, but only poorly on laminin-8. In the case of

laminin-2, mice with a disrupted gene encoding the laminin-2 subunit g1 had

severe peripheral nerve motor deficits and tremor [104]. Disrupting this single

subunit prevented normal elaboration of laminin-2 and a continuous basement

membrane around SCs did not form. SCs in mice without g1 failed to differen-

tiate and synthesize myelin proteins and did not subsequently sort and myeli-

nate axons. Large bundles of naked, unsorted axons were identified and

regeneration was impaired.

Integrins are receptors for extracellular BM molecules, particularly fibronectin

and laminin. On SCs, the major laminin receptors are a-dystroglycan and a6b4

integrin. The nomenclature is confusing since both the laminin ligands and the

integrin receptors have subunits designated by Greek symbols. Other integrins

expressed by SCs include a1b1, a2b1, a5b1, a6b1, avb3, avb1, and b8 [475].

Neurons express a3b1, a4b1, and a5b1 in proprioceptive afferents and a7b1 in

cutaneous afferents [240]. The integrin a1b1 is upregulated in SCs distal to an

axonal crush or transection, including those that did not previously express it

[644]. CD9 is an SC membrane protein associated with integrins and involved in

BM adhesion, proliferation, and migration [18,475].

Integrins are involved in more than adhesive binding alone. They interact

with a large cascade of molecules to effect intracellular signaling (see review [61])

including RHO GTPases, GSKb, PI3K, and a series of “platform” intracellular

molecules that bind the actin cytoskeleton (ILK, a actinin, talin, filamin, vincu-

lin, Arp 2/3, a parvin, and b parvin). When a ligand binds to an integrin, the

receptors become clustered and they recruit actin and its associated proteins to

its cytoplasmic side [61,292].

There are several examples of how interactions between integrin receptors on

neurons, and their extracellular matrix is critical to regeneration. For example,

a7 integrin expressed by injured medium–large diameter sensory neurons

is required to facilitate the actions of NGF and NT-3 in supporting axon out-

growth [210]. a7 and b1 subunits of integrin were not simply expressed in

neurons, but preferentially localized to axons and growth cones of regenerating

facial nerves where they were capable of influencing local behavior. Transgenic

mice lacking a7 had a reduction in axon elongation [736]. Uninjured DRG

sensory neurons have a low level expression of integrin subunit a6 but more

widespread expression of subunits a7 and b1. mRNAs of all three of these

integrin subunits, however, were upregulated after a peripheral axotomy and

all three were expressed in outgrowing axons [717]. This action was specific for

peripheral processes of sensory neurons, since a dorsal root axotomy did not

alter their expression. Sensory neurons thus have the capability of distributing


their integrin receptors to different parts of their structure and it is possible

that selective distribution alters how processes respond to injury. In addition,

blocking the b1 integrin receptor in the above work interrupted SC adhesion.

These findings support the probability that the combination of BM and integrin

are decisive in guiding the specificity of axon regeneration. During development,

proprioceptive neurons in vitro that expressed a3b1, a4b1, and a5b1 extended

more robustly on laminin and fibronectin, whereas cutaneous neurons expressed

higher levels of a7b1 and preferred outgrowth along laminin [240]. While not

yet demonstrated, it may be that regeneration can recapitulate these specific

interactions that occur during development.

Laminin or other BM–integrin interactions thus influence neurite outgrowth

in several ways (Figure 10.2). Point contacts, analogous to focal adhesions of

fibroblasts, are critical to growth cone advance in neurons. Their adhesion to the

basement membrane allows the force generation for growth cone movement.

Figure 10.2 Immunohistochemical image of outgrowing axons (labeled with an

anti-bIII tubulin antibody, red) at the regenerative front of a transected rat sciatic

nerve (connected by a conduit) closely associated with deposits of laminin

(labeled with an anti-laminin antibody, green), likely laid down by leading SCs.

(Bar ¼ 50 microns) (Image taken by Chu Cheng, Zochodne laboratory and

reproduced with permission from [450].) See color plate section.


Downstream of integrin activation by laminin or fibronectin, there are roles for

RHO family GTPases, discussed in Chapter 5. RAC1, a member of this family,

promotes initial point contact after integrin activation. Subsequent stabilization

of these contacts requires RHOA/ROK participation. In keeping with RHOA/ROK’s

inhibitory actions, stabilization implies temporary growth arrest. Semaphorins

that collapse growth cones appear capable of disrupting point contact adhesion

by inhibiting RAC1 [747].

Laminin is therefore a common facilitator for nerve regeneration, its support

being synergistic with the specific growth factors required for outgrowth. Thus,

extracellular BM molecules can substantially influence how neurons respond

to growth factors. For example, embryonic Bax-/- mice grown on laminin acceler-

ated their growth in response to exogenous NGF. On another substrate L1, there

was recruitment of endogenous RHOA within growth cones that prevented

enhanced outgrowth by NGF. Inhibition of RHOA restored the responsiveness

and permitted NGF to accelerate outgrowth on L1 [411]. Large sensory neurons

and CGRP expressing sensory neurons underwent considerable neurite out-

growth on laminin alone and upregulated their expression of the integrin recep-

tor subunit b1 [698]. Both types of neurons had yet further outgrowth when NGF

was also applied. Outgrowth also depended on activation of the PI3K-Akt and

MEK/MAPK, essential growth factor signaling pathways discussed in previous

chapters. In contrast, IB4 NGF-insensitive TrkA negative neurons were more

fastidious in their response to growth conditions. These neurons only extended

neurites when laminin was combined with added GDNF.

Adding laminin and collagen Type 1 to collagen gel preparations stimulates

axon outgrowth from explants [685]. The delivery of exogenous laminin to

regeneration conduits spanning rat sciatic nerve transections enhanced

regrowth in vivo [103]. Laminin was incorporated into the regenerative bridge

material. This valuable support, however, does not appear to be offered by

vascular laminin despite its location near regrowing axons [450].

F-Spondin is an extracellular basement membrane adhesive protein, synthe-

sized by SCs, that is massively upregulated in nerve trunks distal to injury sites

[82]. It is composed of reelin, spondin, and thrombospondin-type 1 repeats (TSR)

domains, each of which may have varying influence on axon regrowth. Interest-

ingly, F-Spondin is less upregulated in sensory than in motor branches.

Other constituents of basmement membranes associated with SCs include

fibronectin, tenascins, Type IV and VI collagen, heparan sulphate, chondroitin

sulphate proteoglycans (see below), and entactin/nidogen [78]. Fibronectin

promotes neurite growth but may have less impact on direction finding and

outgrowth than laminin [243]. Tenascins (TNs) are extracellular basement

membrane glycoprotein “recognition molecules” [244]. Several family members


are described including tenascin-C (TNC), restrictin/J1-160/180 (tenascin-R,TNR),

tenascin-X (TNX), tenascin-Y (TNY), and tenascin-W (TNW). TNC was originally

linked to inhibition of fibronectin adhesion and decreased neurite outgrowth

[114,445]. Subsequent studies have suggested TNC promotes in vivo axon out-

growth, since TNC-deficient mice have delayed reinnervation of their vibrissae.

In contrast, TNR null mice experience better recovery [244]. TNC is normally found

around nodes of Ranvier, is associated with SCs, and is found in the perineurium

[439]. Overall, TN (probably mainly TNC) is upregulated in SC basement mem-

branes after injury and found in proximal nerve stumps, regenerating bridges,

and distal stumps after a nerve transection injury. SCs elaborate TN, and TNC

helps support regeneration by fostering axonal regrowth [439]. In contrast, TNR

may decline in injured peripheral nerves [244]. TNR, however, is also expressed

in perineuronal nets around motor neurons and it protects motor neurons

from microglial-induced cell death [17,244]. TNY is found in rootlets and TNX is

particularly widespread on perineurium and endoneurium [445].

Chrondroitin sulphate proteoglycans (CSPGs) inhibit axon growth in peripheral

neurons. They are concentrated in a narrow band surrounding axon–SC units

and within nodes of Ranvier in normal nerves. Importantly, however, CSPGs are

substantially upregulated (sevenfold in one investigation) in the distal stumps of

peripheral nerves after injury and create an inhibitory barrier for regrowth [812].

CSPG expression in SC BMs becomes thicker and more diffuse throughout the

endoneurium. While CSPGs are capable of activating RHO–ROK in adult sensory

neurons, they also interfere with integrin signaling as occurs during growth on

laminin [780]. Thus CSPGs can disrupt the facilitation of axon regrowth normally

associated with laminin substrates [811]. Their inhibition, in turn, can be over-

come in preconditioned neurons with NGF that acts through PI3K to inhibit

GSK3b. Other myelin inhibitors, such as Nogo, can be overcome by integrin

signaling from laminin or RHO–ROK inhibition. Overall then, it is apparent

that integrins, RHO–ROK, growth factors, and CSPGs are linked in complex

hierarchies that influence outgrowth. To prevent inhibition of axon outgrowth

by CSPGs, however, there is also another innovative approach that involves

matrix metalloproteinases, discussed next.

Matrix metalloproteinases (MMPs) can degrade inhibitory CSPGs and facilitate

regeneration. For example, MMPs promote adult sensory neurite outgrowth

along the basal laminae of peripheral nerves or the regrowth of peripheral

nerves into acellular nerve grafts [180,362]. These roles are exemplified by

MMP-2 and MMP-9, members of the MMP family. MMP-2 is expressed in sensory

neurons and is transported to growth cones [811]. Like the integrins discussed

above, MMPs can be directed intracellularly within neurons to facilitate where

needed. Muir and colleagues labeled this unique feature of growth cones as


“focalized” proteolysis. In a similar approach, transplanted stem cells expressing

MMP-2 increased axon regrowth into chronically denervated nerves by degrading

CSPGs [258]. After crush injury, MMP-2 and MMP-9 levels also rise in distal nerve

stumps despite the absence of axons, a property providing a more permissive

microenvironment within partly degenerated distal nerve stumps [180]. The

findings also indicate that MMPs are not solely expressed in axons, but are also

upregulated in distal nerve stumps by other cell types.

A serpin protease known as nexin-1 may also facilitate regeneration. In the

report by Lino et al. [409], SC expression of nexin-1 had diverse effects on periph-

eral nerve including control of fibrin deposition, tissue plasminogen activator

(tPA) activity, production of mature BDNF, and breakdown of p75. Nexin-1 is

thought to inhibit serine proteases including thrombin, trypsin, plasmin, tPA,

and uPA. While it is thought to promote SC proliferation and survival, its overall

impact may be complex. For example, tPA, one of the substrates it inhibits, is also

required to promote nerve regeneration [624,625].

Further molecules and pathways that influence regeneration

The number of additional pathways and molecules that impact nerve

regeneration is extensive. The evidence for the importance of any particular

molecule varies. Several hormones including estrogen, testosterone, thyroxine,

growth hormone, and ACTH are reported to enhance nerve regeneration

[116,610,650,712]. Forskolin is thought to increase regeneration by activating

adenyl cyclase and increasing cAMP levels [339,340,348].

Cell surface gangliosides are sialic acid-containing glycosphingolipids expressed

on distal portions of the axon and other sites within peripheral nerves. They

include GM1–GM3, GD1–GD3, and GTs based on the number of sialic acid

residues they possess (GM ¼ one; GD ¼ two; GT ¼ 3). There is evidence that GD1a

and GT1b facilitate regeneration. When a specific antibody to GD1a ganglioside

was passively administered to mice, there was a severe inhibition of regeneration

[385]. Abortive “dystrophic” axon swellings (similar to axon endbulbs described

in Chapter 3) developed. These findings are relevant to human neuropathies.

Antibodies to gangliosides have been detected in patients with immune mediated

neuropathies such as Guillain–Barre syndrome. Some antibodies are primarily

directed toward axons and others toward myelin or SCs.

Immunophilins (IPs) are a family of receptor proteins interacting with ligands

that are specific immunosuppressive compounds. IPs were discovered to increase

axon regeneration [223–225,500,637,700,721]. The first IP ligands studied in the

nervous system were cyclosporin A acting on cyclophilin receptors and FK506

(tacrolimus) acting on FK506-binding proteins (FKBPs). Both cyclosporin A and

213Further molecules and pathways that influence regeneration

tacrolimus are potent immunosuppressive agents used in oncology and organ

transplantation. Rapamycin is an additional ligand that binds to a FKBP. IPs

ligate FKBP-12, an interaction that mediates their immunosuppressive actions.

They also ligate FKBP-52, a receptor thought directly responsible for accelerating

nerve regeneration. Non-FKBP-12 binding compounds, including several novel

ligands discovered by Gold and colleagues influence nerve regeneration but do

not exert immunsuppressive actions [222,560]. Ligand–FKBP-12 interactions sup-

press transplant rejection largely because of their ability to inhibit calcineurin

phophatase in T-lymphocytes. In neurons, their action may not involve calci-

neurin but instead involves the regulation of calcium flux through ryanodine

and inositol 1,4,5-triphosphate (IP3) receptors. Additional downstream pathways

include GAP43/B50, Ras/Raf/MAPK, PLC, and PI3K [560,637]. IP ligands increase

the number and maturity of regenerating axons, increase collateral sprouting,

collaborate with NGF, and may accelerate Wallerian degeneration. They also

offer neuroprotection. FK-506 doubled the number of axons and increased

the myelin thickness of peripheral axons extending past an injury zone [659].

Some of the machinery for IP signaling, specifically FKBP-12, is upregulated in

regenerating neurons [590]. Since both animal models and human patients have

benefited from the actions of IPs, they may be valuable agents in situations

requiring combined immunosuppression and nerve grafting (e.g., foreign nerve

grafts to bridge nerve transections) [637]. Novel IP ligands without immunosup-

pressive actions may have direct therapeutic roles in nerve injury.

Phospholipase A2 (PLA2) is a member of a family of enzymes that hydrolyze

phospholipids to form lysophospholipids and arachidonic acid [685]. PLA2 and

arachidonic acid may open calcium channels in conjunction with FGFs and CAMs

to stimulate neurite outgrowth [718]. In a frog sciatic nerve explant assay, PLA2

inhibitors attenuated nerve regeneration, whereas stimulating this pathway

increased regeneration [685].

Semaphorins and their receptors, the neuropilins (NPs), belong to a family of

molecules best known for providing inhibitory guidance cues during develop-

ment. Levels of Sema3A and Sema3F, along with their receptors NP-1 and NP-2,

increase in peripheral nerves distal to crush or transection injuries [602].

NP levels decline between days 8 and 12 after crush but increase on days 24

and 58, coinciding with increased axon ingrowth into the distal stump. SCs

express NP-2 mRNA, whereas Sema3F is found in epineurial fibroblasts and in

the perineurium. The function and specific cellular localization of all of the

semphorin subtypes and their receptors during peripheral nerve regeneration

are not yet known. It is likely, however, that their expression may be important

in fasciculation and direction finding through Bands of Bungner. Inhibitory cues

guide axons in specific trajectories during development.


Another guidance molecule recently identified within peripheral neurons is

“repulsive guidance molecule” (RGM), a glycosylphosphatidylinositol-anchored

protein involved in the collapse of inappropriately projecting embryonic retino-

tectal axons [446]. RGM increases in ventral horn neurons after distal axotomy,

but may have a potential role as an inhibitor of nerve regeneration, an avenue

for exploration [150].

Peptides. Transected peripheral nerve axons accumulate neuropeptides in

their endbulbs that may influence regenerative events (see Chapter 3) [784,788].

For example, vasoactive intestinal polypeptide (VIP), expressed prominently in

axonal endbulbs, applied between the proximal and distal stump of a mouse

sciatic nerve transection gap increased the number and maturity of bridging

axons [773]. CGRP is upregulated in newly regenerating outgrowing peripheral

nerve axons. Moreover, rises in local axonal CGRP expression contrast with its

simultaneous downregulation in perikarya [401]. As endbulbs break down, CGRP

egresses from endbulbs into the extracellular space where it may have several

actions, including vasodilation of local microvessels as discussed in previous

chapters. Little evidence to date, however, has emerged that early CGRP accumu-

lation, persistent expression, or extracellular delivery might have a meaningful

impact on subsequent axonal regeneration. Toth and colleagues [691], however,

noted that siRNAs to both CGRP and a component of its receptor complex RAMP-1

administered in a regeneration conduit spanning rat sciatic nerve transections

completely interrupted nerve regeneration. Moreover, CGRP mRNA was identi-

fied within regenerating axons indicating local synthesis of the peptide. These

findings have suggested that CGRP elaborated by regenerating axons may offer

essential signals to their SC regenerating partners to facilitate regrowth.

Age

Age alters the capacity of the peripheral nervous system to regenerate.

Both regenerative and collateral axon sprouting are attenuated by advanced age.

The topic has been reviewed in depth (see [337,706]). Almost all facets of periph-

eral nerve biology change with age. Intact peripheral nerves undergo reductions

in conduction velocity and myelin thickness. There are declines in the number

of unmyelinated and myelinated axons. The number of sensory receptors as well

as sweat gland innervation diminish. Muscle power and autonomic responses

diminish. There is axonal atrophy, lower levels of myelin proteins, segmental

demyelination, slowing of axoplasmic transport, and decreases in nerve blood

flow. There are rises in the thresholds to sensation.

Increased age slows the progress of Wallerian degeneration through impaired

macrophage recruitment and decreased SC proliferation. The rate of subsequent

215Age

axon regeneration including the numbers of regenerating axons is also attenuated

by aging. Decreased levels of trophic factors made by SCs and decreased expres-

sion of trophic factor receptors contribute to the deficits. Remodeling of the

neuromuscular junction after muscle fiber reinnervation proceeds more slowly,

and reinnervating axons have a delay in their myelination. Retrograde neuronal

changes are also more pronounced with aging. Finally, cutaneous collateral

reinnervation and collateral reinnervation of sweat glands are impaired

[138,499].

Diabetes mellitus and regeneration

Diabetes mellitus, the most common cause of acquired disease of

peripheral nerves, poses major barriers to successful regeneration. There is a

“double hit” in diabetes because neurons and axons spontaneously degenerate

during the disease, yet attempts at regeneration are severely compromised.

Recent reviews on the topic are available [336, 762]. Several steps during the

regenerative process are abnormal in diabetes: axon sprouting [46], elongation

[166], interaction with basement membrane adhesion molecules, and either

upregulation of RAGs (regeneration associated genes) or downregulation of other

genes [46,160,166,300,317,353,413,414,434,448,617,671,672,692,751,752,756,763].

Macrophage entry is slowed in crushed and transected diabetic sciatic nerves

but once influx occurs, their clearance from the nerve is delayed [317,332,671].

Slowed Wallerian-like degeneration with a failure to provide the appropriate

microenvironment for regrowth is associated with impaired regeneration.

Microangiopathy, disease of vasa nervorum from diabetes may contribute

toward an unfavorable microenvironment for regrowth. Rises in blood flow

and growth of new blood vessels after a nerve trunk injury are attenuated in

diabetes [333]. Diabetes is associated with a series of metabolic abnormalities,

each of which may also impact regeneration: oxidative and nitrergic stress,

excessive polyol (sugar alcohol) flux through the conversion of glucose into

sorbitol, changes in PKC subunits, and defective availability or uptake of growth

factors including IGF-1 and insulin. Retrograde loss of parent perikarya may be

excessive, reducing the pool of available parent axons for regeneration. Schwann

cells that normally provide support for axons may be limited in their ability

to proliferate and to synthesize growth factors for axons. Finally, specific alter-

ations of the adhesive properties of basement membrane proteins may influence

regrowth in diabetes.

Overall, diabetes mellitus targets almost every step of the regenerative pro-

cess. Thus, while interventions may be discovered to arrest degenerative diabetic

neuropathy, return to function through regenerative activity may be limited.


Unfortunately, diabetic subjects are also prone to focal neuropathies such as

carpal tunnel syndrome.

Growth of axons from peripheral neurons in the CNS

Peripheral neurons must navigate portions of the CNS. This occurs when

motor axons travel from the ventral horn of the spinal cord into the ventral root,

or when the “central” branches of primary sensory neurons enter the dorsal root

entry zone (DREZ) of the spinal cord. Large myelinated sensory afferents enter

into the posterior columns of gracilis and cuneatus to project to more rostral

nuclei. Other sensory afferents enter dorsal horn gray matter to synapse on

laminae containing interneurons, projection neurons, or motor neurons. For

peripheral neurons, navigation within the CNS poses daunting challenges. The

CNS microenvironment is more hostile to their regrowth than the periphery.

This important topic is reviewed elsewhere [297].

Steinmetz and colleagues [639] facilitated the entry of several afferent

branches into the spinal cord by combining preconditioning with local injec-

tions near the DREZ of chondroitinase ABC to digest inhibitory proteoglycans.

Simultaneous zymosan was injected into the DRG to generate a “proregenera-

tive” inflammatory state. Neumann et al. [504] described enhanced regeneration

of sensory neuron central branches into spinal cord white matter when two

peripheral branch conditioning (priming) lesions were carried out. The first

priming lesion was produced by transection of the sciatic nerve at the same time

as an injury to the dorsal column of the spinal cord. The second sciatic priming

lesion was given 1 week later. The authors hypothesized that the first lesion

enhanced the intrinsic growth capacity of the regenerating central branch

sensory axons, and that the second lesion helped to sustain regrowth. Hoang

et al. [273] demonstrated functional reinnervation of the rat lower urinary tract

by spinal cord motor neurons after reimplantation of avulsed L6 and S1 ventral

roots into the conus medullaris. Wong et al. [746] facilitated the regrowth of

central branch sensory axons past a root crush zone and into the dorsal root

entry zone by using a viral transfection approach. Local injections of an equine

infectious anemia virus (EIAV)-based lentiviral vector expressing the retinoic acid

receptor, RARb2, were made. The injections were associated with functional

improvement and with rises in sensory neuron levels of cAMP. Qui et al. [564]

demonstrated that rises in cAMP, initially mimicking a conditioning lesion oper-

ate through PKA to overcome inhibition by MAG/myelin in the CNS. DRG injection

of db-cAMP, a stable analog of cAMP, facilitated entry of central branches of

sensory neurons into the dorsal columns. Upregulation of arginase I, an enzyme

involved in polyamine synthesis was a further result of cAMP facilitation [83].

217Growth of axons from peripheral neurons in the CNS

Polyamines, in turn are linked to the generation of new rounds of transcription

that may support new outgrowth. Thus this pathway may have facilitated regen-

eration of peripheral axons into the spinal cord by reversing myelin/MAG inhib-

ition or by directly altering growth cone cytoskeletal dynamics.

A number of interventions, including conditioning, cAMP, digestion of PSGPs,

or other approaches, thus have prompted some ingrowth of axons into the CNS.

These successes, however, have had a modest impact at best. Clinical situations

requiring regrowth of axons into dorsal columns develop less frequently than

peripheral injuries.

Conditioning lesions

A conditioning lesion is an intervention thought to “prime” axons for

faster regrowth. It involves a deliberate preliminary injury to an axon followed

by a “test” injury [187,460,461,518,610]. Classically, conditioning lesions are

placed further distally along the nerve beyond the site of the test injury. Condi-

tioning accelerates the outgrowth of neurites in vitro or axon outgrowth in vivo.

It is successful if carried out between 3 days and 2 weeks prior to the “test”

injury, and its onset is more rapid if positioned closer to the cell body.

Conditioning is an interesting and robust phenomenon; several mechanisms

have been considered. These include rises in cAMP levels, enhanced signaling

by integrins, interleukin-6 upregulation, and others [92,167]. Perhaps the

most important mechanism in generating a conditioning response is the early

induction of RAGs in perikarya, considered in Chapter 5. This shift in gene

expression prepares the neuron for regeneration by upregulating growth associ-

ated proteins such as GAP43/B50 and HSP27. While conditioning paradigms

are of interest in understanding strategies to accelerate regeneration, they are

obviously not helpful in treating human nerve damage.

The role of activity

Rehabilitative exercise and activity are assumed to facilitate recovery

from peripheral nerve injury. Rehabilitation is crucial in rebuilding muscle

mass and preventing complications of recumbency. Experimental work however,

suggests that activity may, in fact, be detrimental toward nerve regeneration.

Tam and Gordon (see review by Tam and Gordon [664]), demonstrated that

activity impairedmuscle reinnervation after partial denervation in rats. The growth

and spread of the perisynaptic terminal SCs, normally offering guidance to

regrowing and sprouting motor axons were diminished by exercise [663]. The

combined impacts of excessive flux of calcium and local release of acetylcholine


on perisynaptic SC behavior may account for their impaired behavior after activity

[663]. In contrast to activity, brief epochs of exogenous electrical stimulation

improved motor and sensory axon regeneration and increased preferential motor

reinnervation, discussed in Chapter 6 [73,74]. This form of activation clearly

differs from the more continuous recruitment that occurs during exercise.

Summary

This chapter hints at a further orchestrated series of interactions that

either support or impair peripheral nerve regeneration. Many novel pathways

are likely to be added to this compilation. Broadly speaking, however, adhesive

interactions between axons, SCs, and BM play a major role during repair and

regrowth. These pathways can be synergistic with those of growth factors.

Approaches to degrade CSPGs, to apply immunophilins, and to delineate peptide

actions during regeneration are promising avenues toward understanding and

treating peripheral nerve damage. The goal is to reiterate a more complete

interactive panoply of molecules influencing axon and SC growth during opti-

mal regenerative conditions. New fundamental work will underlie the develop-

ment of therapeutic approaches for severe peripheral nerve injuries, injuries

involving the DREZ or challenging conditions such as diabetes or increased age.

Suggested reading

Colman, D. R. & Filbin, M. (2006). Cell adhesion molecules. In Basic Neurochemistry,

ed. G. J. Siegel, R. W. Albers, S. T. Brady, & D. L. Price. Burlington: Academic

Press, pp. 111–122 [121].

Mirsky, R. & Jessen, K. R. (1999). The neurobiology of Schwann cells. Brain Pathology,

9 (2), 293–311 [475].

219Summary

References

1. The Nerve Growth Cone (1991). ed. P. C. Letourneau, S. B. Kater, & E. R. Macagno. New York:

Raven Press.

2. Abe, I., Ochiai, N., Ichimura, H. et al. (2004). Internodes can nearly double in length

with gradual elongation of the adult rat sciatic nerve. J. Orthop. Res., 22, 571–577.

3. Adams, W. E. (1942). The blood supply of nerves. I. Historical review. J. Anat., 76, 323–341.

4. Ahmed, Z., Brown, R. A., Ljungberg, C., Wiberg, M., & Terenghi, G. (1999). Nerve growth

factor enhances peripheral nerve regeneration in non-human primates. Scand. J. Plast.

Reconstr. Surg. Hand Surg., 33, 393–401.

5. Ahmed, Z., Dent, R. G., Suggate, E. L. et al. (2005). Disinhibition of neurotrophin-

induced dorsal root ganglion cell neurite outgrowth on CNS myelin by siRNA-

mediated knockdown of NgR, p75NTR and Rho-A. Mol. Cell Neurosci., 28, 509–523.

6. Ahmed, Z., Suggate, E. L., Brown, E. R. et al. (2006). Schwann cell-derived factor-induced

modulation of the NgR/p75NTR/EGFR axis disinhibits axon growth through CNS

myelin in vivo and in vitro. Brain, 129, 1517–1533.

7. Al-Majed, A. A., Neumann, C. M., Brushart, T. M., & Gordon, T. (2000). Brief electrical

stimulation promotes the speed and accuracy of motor axonal regeneration.

J. Neurosci., 20, 2602–2608.

8. Alabed, Y. Z., Pool, M., Tone, S. O., & Fournier, A. E. (2007). Identification of CRMP4 as

a convergent regulator of axon outgrowth inhibition. J. Neurosci., 27, 1702–1711.

9. Alluin, O., Feron, F., Desouches, C. et al. (2006). Metabosensitive afferent fiber responses

after peripheral nerve injury and transplantation of an acellular muscle graft in

association with schwann cells. J. Neurotrauma, 23, 1883–1894.

10. Aloyo, V. J., Zwiers, H., & Gispen, W. H. (1982). B-50 protein kinase and kinase C in rat

brain. Prog. Brain Res., 56, 303–315.

11. Althini, S., Usoskin, D., Kylberg, A., Kaplan, P. L., & Ebendal, T. (2004). Blocked MAP

kinase activity selectively enhances neurotrophic growth responses. Mol. Cell Neurosci.,

25, 345–354.

12. Amir, R., Michaelis, M., & Devor, M. (2000). Ectopic discharge in primary sensory

neurons depends on intrinsic membrane potential oscillations. In Proceedings of the

9th World Congress on Pain, ed. M. Devor, M. C. Rowbotham, & Z. Wiesenfeld-Hallin,

Seattle: IASP Press, pp. 93–100.

220

13. Amoh, Y., Li, L., Campillo, R. et al. (2005). Implanted hair follicle stem cells form

Schwann cells that support repair of severed peripheral nerves. Proc. Natl. Acad. Sci.

USA, 102, 17734–17738.

14. Anderson, K. D., Merhege, M. A., Morin, M., Bolognani, F., & Perrone-Bizzozero, N. I.

(2003). Increased expression and localization of the RNA-binding protein HuD and

GAP-43 mRNA to cytoplasmic granules in DRG neurons during nerve regeneration.

Exp. Neurol., 183, 100–108.

15. Angeletti, R. H. & Bradshaw, R. A. (1971). Nerve growth factor from mouse

submaxillary gland: amino acid sequence. Proc. Natl. Acad. Sci. USA, 68, 2417–2420.

16. Angeletti, R. H., Bradshaw, R. A., & Wade, R. D. (1971). Subunit structure and amino

acid composition of mouse submaxillary gland nerve growth factor. Biochemistry,

10, 463–469.

17. Angelov, D. N., Walther, M., Streppel, M. et al. (1998). Tenascin-R is antiadhesive for

activated microglia that induce downregulation of the protein after peripheral nerve

injury: a new role in neuronal protection. J. Neurosci., 18, 6218–6229.

18. Anton, E. S., Hadjiargyrou, M., Patterson, P. H., & Matthew, W. D. (1995). CD9 plays

a role in Schwann cell migration in vitro. J. Neurosci., 15, 584–595.

19. Anton, E. S., Weskamp, G., Reichardt, L. F., & Matthew, W. D. (1994). Nerve growth

factor and its low-affinity receptor promote Schwann cell migration. Proc. Natl. Acad.

Sci. USA, 91, 2795–2799.

20. Appenzeller, O., Dhital, K. K., Cowen, T., & Burnstock, G. (1984). The nerves to blood

vessels supplying blood to nerves: the innervation of vasa nervorum. Brain Res., 304,

383–386.

21. Araki, T., Sasaki, Y., & Milbrandt, J. (2004). Increased nuclear NAD biosynthesis and

SIRT1 activation prevent axonal degeneration. Science, 5686, 1010–1013.

22. Arevalo, M. A. & Rodriguez-Tebar, A. (2006). Activation of casein kinase II and

inhibition of phosphatase and tensin homologue deleted on chromosome

10 phosphatase by nerve growth factor/p75NTR inhibit glycogen synthase kinase-3beta

and stimulate axonal growth. Mol. Biol. Cell., 17, 3369–3377.

23. Armati, P. J. (2007). The Biology of Schwann Cells, ed. P. J. Armati, Cambridge, UK:

Cambridge University Press.

24. Arvidson, B. (1979). Distribution of protein tracers in peripheral ganglia. A light and

electron microscopic study in rodents after various modes of tracer administration.

Acta Univ. Ups., 344, 1–72.

25. Asbury, A. K. & Fields, H. L. (1984). Pain due to peripheral nerve damage: an hypothesis.

Neurology, 34, 1587–1590.

26. Atanasoski, S., Boller, D., De Ventura L. et al. (2006). Cell cycle inhibitors p21 and p16

are required for the regulation of Schwann cell proliferation. Glia, 53, 147–157.

27. Atanasoski, S., Scherer, S. S., Sirkowski, E. et al. (2006). ErbB2 signaling in Schwann cells

is mostly dispensable for maintenance of myelinated peripheral nerves and

proliferation of adult Schwann cells after injury. J. Neurosci., 26, 2124–2131.

28. Averill, S., Michael, G. J., Shortland, P. J. et al. (2004). NGF and GDNF ameliorate the

increase in ATF3 expression which occurs in dorsal root ganglion cells in response to

peripheral nerve injury. Eur. J. Neurosci., 19, 1437–1445.

221References

29. Bajestan, S. N., Umehara, F., Shirahama, Y. et al. (2006). Desert hedgehog-patched 2

expression in peripheral nerves during Wallerian degeneration and regeneration.

J. Neurobiol., 66, 243–255.

30. Baloh, R. H., Enomoto, H., Johnson, E. M., Jr., & Milbrandt, J. (2000). The GDNF family

ligands and receptors – implications for neural development. Curr. Opin. Neurobiol., 10,

103–110.

31. Barde, Y. A., Edgar, D., & Thoenen, H. (1982). Purification of a new neurotrophic factor

from mammalian brain. EMBO, 1, 549–553.

32. Barker, A. R., Rosson, G. D., & Dellon, A. L. (2006). Wound healing in denervated tissue.

Ann. Plast. Surg., 57, 339–342.

33. Barr, M. L. & Kiernan, J. A. (1983). The Human Nervous System, 4th edn. Philadelphia:

Harper and Row.

34. Bausen, M., Fuhrmann, J. C., Betz, H., & O’sullivan, G. A. (2006). The state of the actin

cytoskeleton determines its association with gephyrin: role of ena/VASP family

members. Mol. Cell Neurosci., 31, 376–386.

35. Belkas, J. S., Munro, C. A., Shoichet, M. S., & Midha, R. (2005). Peripheral nerve

regeneration through a synthetic hydrogel nerve tube. Restor. Neurol. Neurosci., 23,

19–29.

36. Belkas, J. S., Shoichet, M. S., & Midha, R. (2004). Peripheral nerve regeneration through

guidance tubes. Neurol. Res., 26, 151–160.

37. Bell, M. A. & Weddell, A. G. (1984). A morphometric study of intrafascicular vessels

of mammalian sciatic nerve. Muscle Nerve, 7, 524–534.

38. Bendszus, M. & Stoll, G. (2003). Caught in the act: in vivo mapping of macrophage

infiltration in nerve injury by magnetic resonance imaging. J. Neurosci., 23,

10892–10896.

39. Benn, S. C., Perrelet, D., Kato, A. C. et al. (2002). Hsp27 upregulation and

phosphorylation is required for injured sensory and motor neuron survival. Neuron,

36, 45–56.

40. Bennett, D. L., Boucher, T. J., Michael, G. J. et al. (2006). Artemin has potent

neurotrophic actions on injured C-fibres. J. Peripher. Nerv. Syst., 11, 330–345.

41. Bennett, G. J. & Xie, Y. K. (1988). A peripheral mononeuropathy in rat that produces

disorders of pain sensation like those seen in man. Pain, 33, 87–107.

42. Berkemeier, L. R., Winslow, J. W., Kaplan, D. R. et al. (1991). Neurotrophin-5: a novel

neurotrophic factor that activates trk and trkB. Neuron, 7, 857–866.

43. Berthold, C.-H., Fraher, J. P., King, R. H. M., & Rydmark, M. (2005). Microscopic anatomy

of the peripheral nervous system. In Peripheral Neuropathy, 4th edn., ed. P. J. Dyck &

P. K. Thomas. Philadelphia: Elsevier Saunders, pp. 35–91.

44. Bertrand, J., Winton, M. J., Rodriguez-Hernandez, N., Campenot, R. B., & McKerracher, L.

(2005). Application of rho antagonist to neuronal cell bodies promotes neurite growth

in compartmented cultures and regeneration of retinal ganglion cell axons in the

optic nerve of adult rats. J. Neurosci., 25, 1113–1121.

45. Bharali, L. A. & Lisney, S. J. (1992). The relationship between unmyelinated afferent

type and neurogenic plasma extravasation in normal and reinnervated rat skin.

Neuroscience, 47, 703–712.

222 References

46. Bisby, M. A. (1980). Axonal transport of labeled protein and regeneration rate in nerves

of streptozocin-diabetic rats. Exp. Neurol., 69, 74–84.

47. Bisby, M. A. (1995). Regeneration of peripheral nervous systems axons. In The Axon:

Structure, Function and Patholphysiology, ed. S. G. Waxman, J. D. Kocsis, & P. K. Stys.

New York: Oxford University Press, pp. 553–578.

48. Bito, H. (2003). Dynamic control of neuronal morphogenesis by rho signaling.

J. Biochem. (Tokyo), 134, 315–319.

49. Blondet, B., Carpentier, G., Lafdil, F., & Courty, J. (2005). Pleiotrophin cellular

localization in nerve regeneration after peripheral nerve injury. J. Histochem. Cytochem.,

53, 971–977.

50. Bohn, M. C. (2004). Motoneurons crave glial cell line-derived neurotrophic factor. Exp.

Neurol., 190, 263–275.

51. Bolsover, S. R. (2005). Calcium signalling in growth cone migration. Cell Calcium,

37, 395–402.

52. Borisoff, J. F., Chan, C. C., Hiebert, G. W. et al. (2003). Suppression of Rho-kinase activity

promotes axonal growth on inhibitory CNS substrates. Mol. Cell Neurosci., 22, 405–416.

53. Bove, G. M. & Light, A. R. (1997). The nervi nervorum, missing link for neuropathic

pain? Pain Forum, 6, 181–190.

54. Bovolenta, P. & Mason, C. (1987). Growth cone morphology varies with position in the

developing mouse visual pathway from retina to first targets. J. Neurosci., 7, 1447–1460.

55. Boyd, J. G. & Gordon, T. (2001). The neurotrophin receptors, trkB and p75, differentially

regulate motor axonal regeneration. J. Neurobiol., 49, 314–325.

56. Boyd, J. G. & Gordon, T. (2002). A dose-dependent facilitation and inhibition of

peripheral nerve regeneration by brain-derived neurotrophic factor. Eur. J. Neurosci.,

15, 613–626.

57. Boyd, J. G. & Gordon, T. (2003). Glial cell line-derived neurotrophic factor and

brain-derived neurotrophic factor sustain the axonal regeneration of chronically

axotomized motoneurons in vivo. Exp. Neurol., 183, 610–619.

58. Boyd, J. G. & Gordon, T. (2003). Neurotrophic factors and their receptors in axonal

regeneration and functional recovery after peripheral nerve injury. Mol. Neurobiol.,

27, 277–324.

59. Bradbury, E. J., Burnstock, G., & McMahon, S. B. (1998). The expression of P2X3

purinoreceptors in sensory neurons: effects of axotomy and glial-derived neurotrophic

factor. Mol. Cell Neurosci., 12, 256–268.

60. Brain, S. D., Williams, T. J., Tippins, J. R., Morris, H. R., & MacIntyre, I. (1985). Calcitonin

gene-related peptide is a potent vasodilator. Nature, 313, 54–56.

61. Brakebusch, C. & Fassler, R. (2003). The integrin-actin connection, an eternal love

affair. EMBO J., 22, 2324–2333.

62. Brannstrom, T., Havton, L., & Kellerth, J. O. (1992). Changes in size and dendritic

arborization patterns of adult cat spinal alpha-motoneurons following permanent

axotomy. J. Comp. Neurol., 318, 439–451.

63. Brannstrom, T., Havton, L., & Kellerth, J. O. (1992). Restorative effects of reinnervation

on the size and dendritic arborization patterns of axotomized cat spinal alpha-

motoneurons. J. Comp. Neurol., 318, 452–461.

223References

64. Brannstrom, T. & Kellerth, J. O. (1998). Changes in synaptology of adult cat spinal

alpha-motoneurons after axotomy. Exp. Brain Res., 118, 1–13.

65. Brannstrom, T. & Kellerth, J. O. (1999). Recovery of synapses in axotomized

adult cat spinal motoneurons after reinnervation into muscle. Exp. Brain Res.,

125, 19–27.

66. Bray, R. C., Fisher, A. W., & Frank, C. B. (1990). Fine vascular anatomy of adult rabbit

knee ligaments. J. Anat., 172, 69–79.

67. Brown, M., Jacobs, T., Eickholt, B. et al. (2004). Alpha2-chimaerin, cyclin-dependent

Kinase 5/p35, and its target collapsin response mediator protein-2 are essential

components in semaphorin 3A-induced growth-cone collapse. J. Neurosci., 24,

8994–9004.

68. Brown, W. F. & Chan, K. M. (1997). Quantitative methods for estimating the number

of motor units in human muscles. Muscle Nerve, Supplement 5, S70–S73.

69. Bruck, W. (1997). The role of macrophages in Wallerian degeneration. Brain Pathol.,

7, 741–752.

70. Brunet, A., Datta, S. R., & Greenberg, M. E. (2001). Transcription-dependent and

-independent control of neuronal survival by the PI3K-Akt signaling pathway. Curr.

Opin. Neurobiol., 11, 297–305.

71. Brushart, T. M. (1988). Preferential reinnervation of motor nerves by regenerating

motor axons. J. Neurosci., 8, 1026–1031.

72. Brushart, T. M. (1993). Motor axons preferentially reinnervate motor pathways.

J. Neurosci., 13, 2730–2738.

73. Brushart, T. M., Hoffman, P. N., Royall, R. M. et al. (2002). Electrical stimulation

promotes motoneuron regeneration without increasing its speed or conditioning

the neuron. J. Neurosci., 22, 6631–6638.

74. Brushart, T. M., Jari, R., Verge, V., Rohde, C., & Gordon, T. (2005). Electrical stimulation

restores the specificity of sensory axon regeneration. Exp. Neurol., 194, 221–229.

75. Brussee, V., Cunningham, F. A., & Zochodne, D. W. (2004). Direct insulin signaling of

neurons reverses diabetic neuropathy. Diabetes, 53, 1824–1830.

76. Buchman, V. L., Sporn, M., & Davies, A. M. (1994). Role of transforming growth factor-

beta isoforms in regulating the expression of nerve growth factor and neurotrophin-3

mRNA levels in embryonic cutaneous cells at different stages of development.

Development, 120, 1621–1629.

77. Bunge, M. B., Wood, P. M., Tynan, L. B., Bates, M. L., & Sanes, J. R. (1989). Perineurium

originates from fibroblasts: demonstration in vitro with a retroviral marker. Science,

243, 229–231.

78. Bunge, R. P. & Bunge, M. B. (1984). Tissue culture observations relating to peripheral

nerve development, regeneration, and disease. In Peripheral Neuropathy, ed. P. J. Dyck

P. K. Thomas, E. Lambert, & R. P. Bunge. Toronto: Saunders, pp. 378–399.

79. Bunge, R. P., Bunge, M. B., & Bates, M. (1989). Movements of the Schwann cell nucleus

implicate progression of the inner (axon-related) Schwann cell process during

myelination. J. Cell Biol., 109, 273–284.

80. Buonanno, A. & Fischbach, G. D. (2001). Neuregulin and ErbB receptor signaling

pathways in the nervous system. Curr. Opin. Neurobiol., 11, 287–296.

224 References

81. Burant, C. F., Lemmon, S. K., Treutelaar, M. K., & Buse, M. G. (1984). Insulin resistance of

denervated rat muscle: a model for impaired receptor-function coupling. Am. J. Physiol.,

247, E657–E666.

82. Burstyn-Cohen, T., Frumkin, A., Xu, Y. T., Scherer, S. S., & Klar, A. (1998). Accumulation

of F-spondin in injured peripheral nerve promotes the outgrowth of sensory axons.

J. Neurosci., 18, 8875–8885.

83. Cai, D., Deng, K., Mellado, W. et al. (2002). Arginase I and polyamines act downstream

from cyclic AMP in overcoming inhibition of axonal growth MAG and myelin in vitro.

Neuron, 35, 711–719.

84. Cai, D., Shen, Y., De, B. M., Tang, S., & Filbin, M. T. (1999). Prior exposure to

neurotrophins blocks inhibition of axonal regeneration by MAG and myelin via

a cAMP-dependent mechanism. Neuron, 22, 89–101.

85. Cameron, N. E. & Cotter, M. A. (2001). Diabetes causes an early reduction in autonomic

ganglion blood flow in rats. J. Diabetes Complications, 15, 198–202.

86. Campana, W. M., Li, X., Dragojlovic, N. et al. (2006). The low-density lipoprotein

receptor-related protein is a pro-survival receptor in Schwann cells: possible

implications in peripheral nerve injury 1195. J. Neurosci., 26, 11197–11207.

87. Campana, W. M. & Myers, R. R. (2001). Erythropoietin and erythropoietin receptors

in the peripheral nervous system: changes after nerve injury. FASEB J., 15, 1804–1806.

88. Campana, W. M. & Myers, R. R. (2003). Exogenous erythropoietin protects against

dorsal root ganglion apoptosis and pain following peripheral nerve injury.

Eur. J. Neurosci., 18, 1497–1506.

89. Campenot, R. B. (1979). Independent control of the local environment of somas and

neurites. Methods Enzymol., 58, 302–307.

90. Campenot, R. B. & MacInnis, B. L. (2004). Retrograde transport of neurotrophins: fact

and function. J. Neurobiol., 58, 217–229.

91. Cao, X. & Shoichet, M. S. (2003). Investigating the synergistic effect of combined

neurotrophic factor concentration gradients to guide axonal growth. Neuroscience,

122, 381–389.

92. Cao, Z., Gao, Y., Bryson, J. B. et al. (2006). The cytokine interleukin-6 is sufficient but not

necessary to mimic the peripheral conditioning lesion effect on axonal growth.

J. Neurosci., 26, 5565–5573.

93. Carmeliet, P. & Storkebaum, E. (2002). Vascular and neuronal effects of VEGF in the

nervous system: implications for neurological disorders. Semin. Cell Dev. Biol., 13, 39–53.

94. Caudy, M. & Bentley, D. (1986). Pioneer growth cone morphologies reveal proximal

increases in substrate affinity within leg segments of grasshopper embryos. J. Neurosci.,

6, 364–379.

95. Chadborn, N. H., Ahmed, A. I., Holt, M. R. et al. (2006). PTEN couples Sema3A signalling

to growth cone collapse. J. Cell Sci., 119, 951–957.

96. Chan, J. R., Watkins, T. A., Cosgaya, J. M. et al. (2004). NGF controls axonal receptivity to

myelination by Schwann cells or oligodendrocytes. Neuron, 43, 183–191.

97. Chan, K. M., Stashuk, D. W., & Brown, W. F. (1998). A longitudinal study of the

pathophysiological changes in single human thenar motor units in amyotrophic

lateral sclerosis. Muscle Nerve, 21, 1714–1723.

225References

98. Chan, S. O., Wong, K. F., Chung, K. Y., & Yung, W. H. (1998). Changes in morphology and

behaviour of retinal growth cones before and after crossing the midline of the mouse

chiasm – a confocal microscopy study. Eur. J. Neurosci., 10, 2511–2522.

99. Chang, K., Ido, Y., LeJeune, W., Williamson, J. R., & Tilton, R. G. (1997). Increased sciatic

nerve blood flow in diabetic rats: assessment by “molecular” vs. particulate

microspheres. Am. J. Physiol., 273, E164–E173.

100. Cheema, S. S., Richards, L. J., Murphy, M., & Bartlett, P. F. (1994). Leukaemia inhibitory

factor rescuesmotoneurones fromaxotomy- induced cell death.NeuroReport, 5, 989–992.

101. Chen, S., Rio, C., Ji, R. R. et al. (2003). Disruption of ErbB receptor signaling in adult non-

myelinating Schwann cells causes progressive sensory loss. Nat. Neurosci., 6, 1186–1193.

102. Chen, Y. S., Chung, S. S., & Chung, S. K. (2005). Noninvasive monitoring of diabetes-

induced cutaneous nerve fiber loss and hypoalgesia in thy1-YFP transgenic mice.

Diabetes, 54, 3112–3118.

103. Chen, Y. Y., McDonald, D., Cheng, C. et al. (2005). Axon and Schwann cell partnership

during nerve regrowth. J. Neuropathol. Exp. Neurol., 64, 613–622.

104. Chen, Z. L. & Strickland, S. (2003). Laminin gamma1 is critical for Schwann cell

differentiation, axon myelination, and regeneration in the peripheral nerve. J. Cell

Biol., 163, 889–899.

105. Cheng, C., Webber, C. A., Wang, J. et al. (2008). Activated RhoA and peripheral axon

regeneration. Exp. Neurol. (in press).

106. Cheng, C. & Zochodne, D. W. (2000). Proliferation, migration and phenotypic changes

in vivo of Schwann cells in the presence of myelinated fibers [abstract]. Soc. Neurosci.

Abs., 26, 612.

107. Cheng, C. & Zochodne, D. W. (2001). Dorsal root ganglia contain populations of

proliferating cells [abstract]. Soc. Neurosci. Abs., 31, 315.

108. Cheng, C. & Zochodne, D. W. (2002). In vivo proliferation, migration & phenotypic

changes of Schwann cells in the presence ofmyelinated fibers. Neuroscience, 115, 321–329.

109. Cheng, H. L., Randolph, A., Yee, D. et al. (1996). Characterization of insulin-like growth

factor-I and its receptor and binding proteins in transected nerves and cultured

Schwann cells. J. Neurochem., 66, 525–536.

110. Cheng, H. L., Russell, J. W., & Feldman, E. L. (1999). IGF-I promotes peripheral nervous

system myelination. Ann. N.Y. Acad. Sci., 883, 124–130.

111. Cheng, L., Khan, M., & Mudge, A. W. (1995). Calcitonin gene-related peptide promotes

Schwann cell proliferation. J. Cell Biol., 129, 789–796.

112. Chernousov, M. A., Rothblum, K., Stahl, R. C. et al. (2006). Glypican-1 and alpha4(V)

collagen are required for Schwann cell myelination. J Neurosci., 26, 508–517.

113. Chin, P. C., Majdzadeh, N., & D’Mello, S. R. (2005). Inhibition of GSK3beta is a common

event in neuroprotection by different survival factors. Brain Res. Mol Brain Res., 137,

193–201.

114. Chiquet-Ehrismann, R. (1995). Inhibition of cell adhesion by anti-adhesive molecules.

Curr. Opin. Cell Biol., 7, 715–719.

115. Chung, K., Kim, H. J., Na, H. S., Park, M. J., & Chung, J. M. (1993). Abnormalities of

sympathetic innervation in the area of an injured peripheral nerve in a rat model

of neuropathic pain. Neurosci. Lett., 162, 85–88.

226 References

116. Cockett, S. A. & Kiernan, J. A. (1973). Acceleration of peripheral nervous regeneration

in the rat by exogenous triiodothyronine. Exp. Neurol., 39, 389–394.

117. Coggeshall, R. E. & Lekan, H. A. (1996). Methods for determining numbers of cells and

synapses: a case for more uniform standards of review. J. Comp. Neurol., 364, 6–15.

118. Coggeshall, R. E., Lekan, H. A., Doubell, T. P., Allchorne, A., & Woolf, C. J. (1997). Central

changes in primary afferent fibers following peripheral nerve lesions. Neuroscience,

77, 1115–1122.

119. Coggins, P. J. & Zwiers, H. (1991). B-50 (GAP-43): biochemistry and functional

neurochemistry of a neuron-specific phosphoprotein. J. Neurochem., 56, 1095–1106.

120. Cole, A. R., Causeret, F., Yadirgi, G. et al. (2006). Distinct priming kinases contribute to

differential regulation of collapsin response mediator proteins by glycogen synthase

kinase-3 in vivo. J. Biol. Chem., 281, 16591–16598.

121. Colman, D. R. & Filbin, M. T. (2006). Cell adhesion molecules. In Basic Neurochemistry,

7th edn., ed. G. J. Siegel, R. W. Albers, S. T. Brady, & D. L. Price. Burlington: Academic

Press, pp. 111–122.

122. Conrad, S., Schluesener, H. J., Trautmann, K. et al. (2005). Prolonged lesional expression

of RhoA and RhoB following spinal cord injury. J. Comp. Neurol., 487, 166–175.

123. Costigan,M.,Mannion, R. J., Kendall, G. et al. (1998). Heat shock protein 27: developmental

regulation and expression after peripheral nerve injury. J. Neurosci., 18, 5891–5900.

124. Court, F. A. & Alvarez, J. (2005). Local regulation of the axonal phenotype, a case of

merotrophism. Biol. Res., 38, 365–374.

125. Craner, M. J., Klein, J. P., Black, J. A., & Waxman, S. G. (2002). Preferential expression

of IGF-I in small DRG neurons and down-regulation following injury. NeuroReport, 13,

1649–1652.

126. Curtis, R., Stewart, H. J., Hall, S. M. et al. (1992). GAP-43 is expressed by nonmyelin-

forming Schwann cells of the peripheral nervous system. J. Cell Biol., 116, 1455–1464.

127. Daniels, R. H. & Bokoch, G. M. (1999). p21-activated protein kinase: a crucial

component of morphological signaling? Trends Biochem. Sci., 24, 350–355.

128. Day, T. J., Lagerlund, T. D., & Low, P. A. (1989). Analysis of H2 clearance curves used to

measure blood flow in rat sciatic nerve. J. Physiol., 414, 35–54.

129. Day, T. J., Schmelzer, J. D., & Low, P. A. (1989). Aortic occlusion and reperfusion and

conduction, blood flow and the blood–nerve barrier of rat sciatic nerve. Exp. Neurol.,

103, 173–178.

130. Day, W. A., Koishi, K., & McLennan, I. S. (2003). Transforming growth factor beta 1 may

regulate the stability of mature myelin sheaths. Exp. Neurol., 184, 857–864.

131. de Medinaceli, L., Freed, W. J., & Wyatt, R. J. (1982). An index of the functional

condition of rat sciatic nerve based on measurements made from walking tracks.

Exp. Neurol., 77, 634–643.

132. Degn, J., Tandrup, T., & Jakobsen, J. (1999). Effect of nerve crush on perikaryal number

and volume of neurons in adult rat dorsal root ganglion. J. Comp. Neurol., 412, 186–192.

133. Dent, E. W. & Gertler, F. B. (2003). Cytoskeletal dynamics and transport in growth cone

motility and axon guidance. Neuron, 40, 209–227.

134. Dergham, P., Ellezam, B., Essagian, C. et al. (2002). Rho signaling pathway targeted to

promote spinal cord repair. J. Neurosci., 22, 6570–6577.

227References

135. Devor, M., Govrin-Lippmann, R., & Angelides, K. (1993). Naþ channel immunolocalization

in peripheral mammalian axons and changes following nerve injury and neuroma

formation. J. Neurosci., 13, 1976–1992.

136. Devor, M., Govrin-Lippmann, R., Frank, I., & Raber, P. (1985). Proliferation of primary

sensory neurons in adult rat dorsal root ganglion and the kinetics of retrograde cell

loss after sciatic nerve section. Somatosens. Res., 3, 139–167.

137. Dhital, K. K. & Appenzeller, O. (1988). Innervation of vasa nervorum. In Nonadrenergic

Innervation of Blood Vessels, ed. G. Burnstock, & S. G. Griffith. Boca Raton, FL. CRC Press,

pp. 191–211.

138. Diamond, J. & Foerster, A. (1992). Recovery of sensory function in skin deprived of its

innervation by lesion of the peripheral nerve. Exp. Neurol., 115, 100–103.

139. Diamond, J., Foerster, A., Holmes, M., & Coughlin, M. (1992). Sensory nerves in adult

rats regenerate and restore sensory function to the skin independently of endogenous

NGF. J. Neurosci., 12, 1467–1476.

140. Diamond, J., Gloster, A., & Kitchener, P. (1992). Regulation of the sensory innervation

of skin: trophic control of collateral sprouting. In Sensory Neurons: Diversity, Development

and Plasticity, ed. S. A. Scott. Oxford: Oxford University Press, pp. 309–332.

141. Diamond, J., Holmes, M., & Coughlin, M. (1992). Endogenous NGF and nerve impulses

regulate the collateral sprouting of sensory axons in the skin of the adult rat.

J. Neurosci., 12, 1454–1466.

142. DiStefano, P. S., Friedman, B., Radziejewski, C. et al. (1992). The neurotrophins BDNF,

NT-3, and NGF display distinct patterns of retrograde axonal transport in peripheral

and central neurons. Neuron, 8, 983–993.

143. Dodge, M. E., Rahimtula, M., & Mearow, K. M. (2002). Factors contributing

to neurotrophin-independent survival of adult sensory neurons. Brain Res.,

953, 144–156.

144. Doherty, P., Williams, E., & Walsh, F. S. (1995). A soluble chimeric form of the

L1 glycoprotein stimulates neurite outgrowth. Neuron, 14, 57–66.

145. Doherty, T. J., Chan, K. M., & Brown, W. F. (2007). Motor neurons, motor units, and

motor unit recruitment. In Neuromuscular Function and Disease, ed. W. F. Brown, C. F.,

Bolton, & M. J. Aminoff. Philadelphia: W. B. Saunders, pp. 247–273.

146. Doherty, T. J., Stashuk, D. W., & Brown, W. F. (1993). Determinants of mean motor unit

size: impact on estimates of motor unit number. Muscle Nerve, 16 (12), 1326–1331.

147. Donaldson, D., Evans, O. B., & Harrison, R. W. (1986). Insulin binding in denervated

muscle. Muscle Nerve, 9, 211–215.

148. Dort, J. C., Zochodne, D., Fan, Y., Prychitko, J., & Peeling, J. (1997). Biochemical changes

in denervated muscle identified by magnetic resonance spectroscopy. J. Otolaryngol.,

26, 368–373.

149. Doucette, R. & Diamond, J. (1987). Normal and precocious sprouting of heat

nociceptors in the skin of adult rats. J. Comp. Neurol., 261, 592–603.

150. Doya, H., Ito, T., Hata, K. et al. (2006). Induction of repulsive guidance molecule in

neurons following sciatic nerve injury. J. Chem. Neuroanat., 32, 74–77.

151. DuBois, D. C. & Max, S. R. (1983). Effect of denervation and reinnervation on oxidation

of 6–14C glucose by rat skeletal muscle homogenates. J. Neurochem., 40, 727–733.

228 References

152. Dubovy, P. & Aldskogius, H. (1996). Degeneration and regeneration of cutaneous

sensory nerve formations. Microsc. Res. Tech., 34, 362–375.

153. Duce, I. R., & Keen, P. (1980). The formation of axonal sprouts in organ culture and

their relationship to sprouting in vivo, Int. Rev. Cytol., 66, 211–256.

154. Dumitru, D. (1996). Single muscle fiber discharges (insertional activity, end-plate

potentials, positive sharp waves, and fibrillation potentials): a unifying proposal.

Muscle Nerve, 19 (2), 221–226.

155. Dyck, P. J., Karnes, J., O’Brien, P. et al. (1984). Spatial pattern of nerve fiber abnormality

indicative of pathologic mechanism. Am. J. Pathol., 117, 225–238.

156. Dyck, P. J., Karnes, J., Sparks, M., & Low, P. A. (1982). The morphometric composition

of myelinated fibres by nerve, level and species related to nerve microenvironment

and ischemia. Electroencephalogr. Clin. Neurophysiol. Suppl., 36, 39–55.

157. Dyck, P. J. & Norell, J. E. (1999). Microvasculitis and ischemia in diabetic lumbosacral

radiculoplexus neuropathy. Neurology, 53, 2113–2121.

158. Dyck, P. J. & O’Brien, P. C. (1999). Quantitative sensation testing in epidemiological and

therapeutic studies of peripheral neuropathy. Muscle Nerve, 22, 659–662.

159. Dyck, P. J. & Thomas, P. K. (2005). Peripheral Neuropathy. Philadelphia: Elsevier

Saunders.

160. Dyck, P. J., Zimmerman, B. R., Vilen, T. H. et al. (1988). Nerve glucose, fructose, sorbitol,

myo-inositol, and fiber degeneration and regeneration in diabetic neuropathy. N. Engl.

J. Med., 319, 542–548.

161. Dyck, P. J., Zimmerman, I. R., Johnson, D. M. et al. (1996). A standard test of heat-pain

responses using CASE IV. J. Neurol. Sci., 136, 54–63.

162. Eberhardt, K. A., Irintchev, A., Al-Majed, A. A. et al. (2006). BDNF/TrkB signaling

regulates HNK-1 carbohydrate expression in regenerating motor nerves and promotes

functional recovery after peripheral nerve repair. Exp. Neurol, 198, 500–510.

163. Edstrom, A. & Ekstrom, P. A. (2003). Role of phosphatidylinositol 3-kinase in neuronal

survival and axonal outgrowth of adult mouse dorsal root ganglia explants. J. Neurosci.

Res., 74, 726–735.

164. Ehlers, M. D. (2004). Deconstructing the axon: Wallerian degeneration and the

ubiquitin-proteasome system. Trends Neurosci., 27, 3–6.

165. Eickholt, B. J., Walsh, F. S., & Doherty, P. (2002). An inactive pool of GSK-3 at the leading

edge of growth cones is implicated in Semaphorin 3A signaling. J. Cell Biol., 157,

211–217.

166. Ekstrom, A. R. & Tomlinson, D. R. (1989). Impaired nerve regeneration in

streptozotocin-diabetic rats. Effects of treatment with an aldose reductase inhibitor.

J. Neurol. Sci., 93, 231–237.

167. Ekstrom, P. A., Mayer, U., Panjwani, A. et al. (2003). Involvement of alpha7beta1

integrin in the conditioning-lesion effect on sensory axon regeneration. Mol. Cell

Neurosci., 22, 383–395.

168. Endo, M., Ohashi, K., & Mizuno, K. (2007). LIM kinase and slingshot are critical for

neurite extension. J. Biol. Chem., 282, 13692–13702.

169. Enerback, L., Olsson, Y., & Sourander, P. (1965). Mast cells in normal and sectioned

peripheral nerve. Z. Zellforsch. Mikrosk. Anat., 66, 596–608.

229References

170. English, A. W., Meador, W., & Carrasco, D. I. (2005). Neurotrophin-4/5 is required for the

early growth of regenerating axons in peripheral nerves. Eur. J. Neurosci., 21, 2624–2634.

171. Erez, H., Malkinson, G., Prager-Khoutorsky, M. et al. (2007). Formation of microtubule-

based traps controls the sorting and concentration of vesicles to restricted sites of

regenerating neurons after axotomy. J. Cell Biol., 176, 497–507.

172. Esper, R. M. & Loeb, J. A. (2004). Rapid axoglial signaling mediated by neuregulin and

neurotrophic factors. J. Neurosci., 24, 6218–6227.

173. Esper, R. M., Pankonin, M. S., & Loeb, J. A. (2006). Neuregulins: versatile growth and

differentiation factors in nervous system development and human disease. Brain Res.

Rev., 51, 161–175.

174. Eswarakumar, V. P., Lax, I., & Schlessinger, J. (2005). Cellular signaling by fibroblast

growth factor receptors. Cytokine Growth Factor Rev., 16, 139–149.

175. Etienne-Manneville, S. & Hall, A. (2002). Rho GTPases in cell biology. Nature, 420,

629–635.

176. Eyer, J. & Peterson, A. (1994). Neurofilament-deficient axons and perikaryal aggregates

in viable transgenic mice expressing a neurofilament-beta-galactosidase fusion

protein. Neuron, 12, 389–405.

177. Fadool, D. A., Tucker, K., Phillips, J. J., & Simmen, J. A. (2000). Brain insulin receptor

causes activity-dependent current suppression in the olfactory bulb through multiple

phosphorylation of Kv1.3. J. Neurophysiol., 83, 2332–2348.

178. Fahnestock, M., Yu, G., & Coughlin, M. D. (2004). ProNGF: a neurotrophic or an

apoptotic molecule? Prog. Brain Res., 146, 101–110.

179. Fenrich, K. & Gordon, T. (2004). Canadian Association of Neuroscience review: axonal

regeneration in the peripheral and central nervous systems – current issues and

advances. Can. J. Neurol. Sci., 31, 142–156.

180. Ferguson, T. A. & Muir, D. (2000). MMP-2 and MMP-9 increase the neurite-promoting

potential of schwann cell basal laminae and are upregulated in degenerated nerve.

Mol. Cell Neurosci., 16, 157–167.

181. Fernandes, K. J. L. & Tetzlaff, W. G. (2000). Gene expression in axotomized neurons:

identifying the intrinsic determinants of axonal growth. In Axonal Regeneration in

the Central Nervous System, ed. N. A. Ingoglia & M. Murray. New York: Marcel Dekker,

pp. 219–266.

182. Fernyhough, P., Smith, D. R., Schapansky, J. et al. (2005). Activation of nuclear factor-

kappaB via endogenous tumor necrosis factor alpha regulates survival of axotomized

adult sensory neurons. J. Neurosci., 25, 1682–1690.

183. Filbin, M. T. (2003). Myelin-associated inhibitors of axonal regeneration in the adult

mammalian CNS. Nat. Rev. Neurosci., 4, 703–713.

184. Filbin, M. T. (2006). How inflammation promotes regeneration. Nat. Neurosci., 9, 715–717.

185. Fischer, D., Petkova, V., Thanos, S., & Benowitz, L. I. (2004). Switching mature retinal

ganglion cells to a robust growth state in vivo: gene expression and synergy with RhoA

inactivation. J. Neurosci., 24, 8726–8740.

186. Folli, F., Bonfanti, L., Renard, E., Kahn, C. R., & Merighi, A. (1994). Insulin receptor

substrate-1 (IRS-1) distribution in the rat central nervous system. J. Neurosci., 14,

6412–6422.

230 References

187. Forman, D. S., McQuarrie, I. G., Labore, F. W. et al. (1980). Time course of the

conditioning lesion effect on axonal regeneration. Brain Res., 182, 180–185.

188. Fournier, A. E., Takizawa, B. T., & Strittmatter, S. M. (2003). Rho kinase inhibition

enhances axonal regeneration in the injured CNS. J. Neurosci., 23, 1416–1423.

189. Franke, T. F., Hornik, C. P., Segev, L., Shostak, G. A., & Sugimoto, C. (2003). PI3K/Akt and

apoptosis: size matters. Oncogene, 22, 8983–8998.

190. Franz, C. K., Rutishauser, U., & Rafuse, V. F. (2005). Polysialylated neural cell adhesion

molecule is necessary for selective targeting of regenerating motor neurons.

J. Neurosci., 25, 2081–2091.

191. Frey, D., Laux, T., Xu, L., Schneider, C., & Caroni, P. (2000). Shared and unique roles of

CAP23 and GAP43 in actin regulation, neurite outgrowth, and anatomical plasticity.

J. Cell Biol., 149, 1443–1454.

192. Fried, K. & Frisen, J. (1990). End structure and neuropeptide immunoreactivity of

axons in sciatic neuromas following nerve section in neonatal rats. Exp. Neurol., 109,

286–293.

193. Fried, K., Govrin-Lippmann, R., Rosenthal, F., Ellisman, M. H., & Devor, M. (1991).

Ultrastructure of afferent axon endings in a neuroma. J. Neurocytol., 20, 682–701.

194. Friede, R. L. & Beuche, W. (1985). A new approach toward analyzing peripheral nerve

fiber populations. I. Variance in sheath thickness corresponds to different geometric

proportions of the internodes. J. Neuropathol. Exp. Neurol., 44, 60–72.

195. Friede, R. L. & Bischhausen, R. (1980). The fine structure of stumps of transected nerve

fibers in subserial sections. J. Neurol. Sci., 44, 181–203.

196. Frost, J. A., Swantek, J. L., Stippec, S. et al. (2000). Stimulation of NFkappa B activity by

multiple signaling pathways requires PAK1. J. Biol. Chem., 275, 19693–19699.

197. Frost, J. A., Xu, S., Hutchison, M. R., Marcus, S., & Cobb, M. H. (1996). Actions of Rho

family small G proteins and p21-activated protein kinases on mitogen-activated

protein kinase family members. Mol. Cell Biol., 16, 3707–3713.

198. Fry, E. J., Ho, C., & David, S. (2007). A role for Nogo receptor in macrophage clearance

from injured peripheral nerve. Neuron, 53, 649–662.

199. Fu, S. Y. & Gordon, T. (1995). Contributing factors to poor functional recovery after

delayed nerve repair: prolonged denervation. J. Neurosci., 15, 3886–3895.

200. Fu, S. Y. & Gordon, T. (1997). The cellular and molecular basis of peripheral nerve

regeneration. Mol. Neurobiol., 14, 67–116.

201. Fukata, Y., Itoh, T. J., Kimura, T. et al. (2002). CRMP-2 binds to tubulin heterodimers to

promote microtubule assembly. Nat. Cell Biol., 4, 583–591.

202. Fukuda, T., Kiuchi, K., & Takahashi, M. (2002). Novel mechanism of regulation of

Rac activity and lamellipodia formation by RET tyrosine kinase. J. Biol. Chem., 277,

19114–19121.

203. Funakoshi, H., Frisen, J., Barbany, G. et al. (1993). Differential expression of mRNAs

for neurotrophins and their receptors after axotomy of the sciatic nerve. J. Cell Biol.,

123, 455–465.

204. Furey, M. J., Midha, R., Xu, Q. G., Belkas, J., & Gordon, T. (2007). Prolonged target

deprivation reduces the capacity of injured motoneurons to regenerate. Neurosurgery,

60, 723–732.

231References

205. Galbraith, J. A. & Gallant, P. E. (2000). Axonal transport of tubulin and actin.

J. Neurocytol., 29, 889–911.

206. Gallo, G. & Letourneau, P. C. (1998). Localized sources of neurotrophins initiate axon

collateral sprouting. J. Neurosci., 18, 5403–5414.

207. Gallo, G. & Letourneau, P. C. (2003). Regulation of growth cone actin filaments by

guidance cues. J. Neurobiol., 58, 92–102.

208. Galtrey, C. M. & Fawcett, J. W. (2007). Characterization of tests of functional recovery

after median and ulnar nerve injury and repair in the rat forelimb. J. Peripher. Nerv.

Syst., 12, 11–27.

209. Garbay, B., Heape, A. M., Sargueil, F., & Cassagne, C. (2000). Myelin synthesis in the

peripheral nervous system. Prog. Neurobiol., 61, 267–304.

210. Gardiner, N. J., Fernyhough, P., Tomlinson, D. R. et al. (2005). Alpha7 integrin mediates

neurite outgrowth of distinct populations of adult sensory neurons. Mol. Cell Neurosci.,

28, 229–240.

211. Garratt, A. N., Voiculescu, O., Topilko, P., Charnay, P., & Birchmeier, C. (2000). A dual

role of erbB2 in myelination and in expansion of the Schwann cell precursor pool.

J. Cell Biol., 148, 1035–1046.

212. Gavazzi, I., Kumar, R. D., McMahon, S. B., & Cohen, J. (1999). Growth responses of

different subpopulations of adult sensory neurons to neurotrophic factors in vitro.

Eur. J. Neurosci., 11, 3405–3414.

213. Gehler, S., Gallo, G., Veien, E., & Letourneau, P. C. (2004). p75 neurotrophin receptor

signaling regulates growth cone filopodial dynamics through modulating RhoA

activity. J. Neurosci., 24, 4363–4372.

214. Ghosh, M., Song, X., Mouneimne, G. et al. (2004). Cofilin promotes actin polymerization

and defines the direction of cell motility. Science, 304, 743–746.

215. Giannini, C. & Dyck, P. J. (1990). The fate of Schwann cell basement membranes in

permanently transected nerves. J. Neuropathol. Exp. Neurol., 49, 550–563.

216. Giehl, K. M. & Tetzlaff, W. (1996). BDNF and NT-3, but not NGF, prevent axotomy-

induced death of rat corticospinal neurons in vivo. Eur. J. Neurosci., 8, 1167–1175.

217. Giniger, E. (2002). How do Rho family GTPases direct axon growth and guidance?

A proposal relating signaling pathways to growth cone mechanics. Differentiation,

70, 385–396.

218. Giraux, P., Sirigu, A., Schneider, F., & Dubernard, J. M. (2001). Cortical reorganization

in motor cortex after graft of both hands. Nat. Neurosci., 4, 691–692.

219. Glazner, G. W. & Ishii, D. N. (1995). Insulinlike growth factor gene expression in rat

muscle during reinnervation. Muscle Nerve, 18, 1433–1442.

220. Glazner, G. W., Morrison, A. E., & Ishii, D. N. (1994). Elevated insulin-like growth factor

(IGF) gene expression in sciatic nerves during IGF-supported nerve regeneration. Brain

Res. Mol. Brain Res., 25, 265–272.

221. Godement, P., Wang, L. C., & Mason, C. A. (1994). Retinal axon divergence in the optic

chiasm: dynamics of growth cone behavior at the midline. J. Neurosci., 14, 7024–7039.

222. Gold, B. G., Armistead, D. M., & Wang, M. S. (2005). Non-FK506-binding protein-12

neuroimmunophilin ligands increase neurite elongation and accelerate nerve

regeneration. J. Neurosci. Res., 80, 56–65.

232 References

223. Gold, B. G., Katoh, K., & Storm-Dickerson, T. (1995). The immunosuppressant FK506

increases the rate of axonal regeneration in rat sciatic nerve. J. Neurosci., 15, 7509–7516.

224. Gold, B. G., Yew, J. Y., & Zeleny-Pooley, M. (1998). The immunosuppressant FK506

increases GAP-43 mRNA levels in axotomized sensory neurons. Neurosci. Lett., 241,

25–28.

225. Gold, B. G., Zeleny-Pooley, M., Wang, M. S., Chaturvedi, P., & Armistead, D. M. (1997).

A nonimmunosuppressant FKBP-12 ligand increases nerve regeneration. Exp. Neurol.,

147, 269–278.

226. Goldberg, D. J., Foley, M. S., Tang, D., & Grabham, P. W. (2000). Recruitment

of the Arp2/3 complex and mena for the stimulation of actin polymerization in

growth cones by nerve growth factor. J. Neurosci. Res., 60, 458–467.

227. Gordon, T. & Fu, S. Y. (1997). Long-term response to nerve injury. Adv. Neurol., 72, 185–199.

228. Gordon-Weeks P. (2000). Neuronal Growth Cones. Cambridge, UK: Cambridge University

Press.

229. Goshima, Y., Nakamura, F., Strittmatter, P., & Strittmatter, S. M. (1995). Collapsin-

induced growth cone collapse mediated by an intracellular protein related to UNC-33.

Nature, 376, 509–514.

230. Gotz, R., Koster, R., Winkler, C. et al. (1994). Neurotrophin-6 is a new member of the

nerve growth factor family. Nature, 372, 266–269.

231. Gratto, K. A. & Verge, V. M. (2003). Neurotrophin-3 down-regulates trkA mRNA, NGF

high-affinity binding sites, and associated phenotype in adult DRG neurons. Eur.

J. Neurosci., 18, 1535–1548.

232. Gravel, C., Gotz, R., Lorrain, A., & Sendtner, M. (1997). Adenoviral gene transfer of

ciliary neurotrophic factor and brain-derived neurotrophic factor leads to long-term

survival of axotomized motor neurons. Nat. Med., 3, 765–770.

233. Griesbeck, O., Parsadanian, A. S., Sendtner, M., & Thoenen, H. (1995). Expression of

neurotrophins in skeletal muscle: quantitative comparison and significance for

motoneuron survival and maintenance of function. J. Neurosci. Res., 42, 21–33.

234. Griffin, J. W., George, E. B., & Chaudhry, V. (1996). Wallerian degeneration in

peripheral nerve disease. Baillieres Clin. Neurol., 5, 65–75.

235. Griffin, J. W., George, E. G., Hsieh, S. T., & Glass, J. D. (1995). Axonal degeneration and

disorders of the axonal cytoskeleton. In The Axon. Structure, Function and Pathophysiology,

ed. S. G. Waxman, J. D. Kocsis, & P. K. Stys. Oxford, UK: Oxford University Press,

pp. 375–390.

236. Griffin, J. W., George, R., & Ho, T. (1993). Macrophage systems in peripheral nerves.

A review. J. Neuropathol. Exp. Neurol., 52, 553–560.

237. Grothe, C., Haastert, K., & Jungnickel, J. (2006). Physiological function and putative

therapeutic impact of the FGF-2 system in peripheral nerve regeneration – lessons

from in vivo studies in mice and rats. Brain Res. Rev., 51, 293–299.

238. Grothe, C., Meisinger, C., & Claus, P. (2001). In vivo expression and localization of the

fibroblast growth factor system in the intact and lesioned rat peripheral nerve and

spinal ganglia. J. Comp. Neurol., 434, 342–357.

239. Groves, M. J., Schanzer, A., Simpson, A. J. et al. (2003). Profile of adult rat sensory neuron

loss, apoptosis and replacement after sciatic nerve crush. J. Neurocytol., 32, 113–122.

233References

240. Guan, W., Puthenveedu, M. A., & Condic, M. L. (2003). Sensory neuron subtypes have

unique substratum preference and receptor expression before target innervation.

J. Neurosci., 23, 1781–1791.

241. Guenard, V., Kleitman, N., Morrissey, T. K., Bunge, R. P., & Aebischer, P. (1992).

Syngeneic Schwann cells derived from adult nerves seeded in semipermeable

guidance channels enhance peripheral nerve regeneration. J. Neurosci., 12, 3310–3320.

242. Guertin, A. D., Zhang, D. P., Mak, K. S., Alberta, J. A., & Kim, H. A. (2005). Microanatomy

of axon/glial signaling during Wallerian degeneration. J. Neurosci., 25, 3478–3487.

243. Gundersen, R. W. (1987). Response of sensory neurites and growth cones to patterned

substrata of laminin and fibronectin in vitro. Dev. Biol., 121, 423–431.

244. Guntinas-Lichius, O., Angelov, D. N., Morellini, F. et al. (2005). Opposite impacts of

tenascin-C and tenascin-R deficiency in mice on the functional outcome of facial nerve

repair. Eur. J. Neurosci., 22, 2171–2179.

245. Gustafsson, H., Tamm, C., & Forsby, A. (2004). Signalling pathways for insulin-like

growth factor type 1-mediated expression of uncoupling protein 3. J. Neurochem.,

88, 462–468.

246. Haapaniemi, T., Nishiura, Y., & Dahlin, L. B. (2002). Functional evaluation after rat

sciatic nerve injury followed by hyperbaric oxygen treatment. J. Peripher. Nerv. Syst.,

7, 149–154.

247. Hall, S. M. (1986). The effect of inhibiting Schwann cell mitosis on the re-innervation of

acellular autografts in the peripheral nervous system of the mouse. Neuropathol. Appl.

Neurobiol., 12, 401–414.

248. Hall, S. M. (1999). The biology of chronically denervated Schwann cells. Ann. N.Y. Acad.

Sci., 883, 215–233.

249. Hamanoue, M., Middleton, G., Wyatt, S. et al. (1999). p75-mediated NF-kappaB

activation enhances the survival response of developing sensory neurons to nerve

growth factor. Mol. Cell Neurosci., 14, 28–40.

250. Hammarberg, H., Piehl, F., Cullheim, S. et al. (1996). GDNF mRNA in Schwann cells and

DRG satellite cells after chronic sciatic nerve injury. NeuroReport, 7, 857–860.

251. Hanada, M., Feng, J., & Hemmings, B. A. (2004). Structure, regulation and function of

PKB/AKT – a major therapeutic target. Biochim. Biophys. Acta., 1697, 3–16.

252. Hanz, S., Perlson, E., Willis, D. et al. (2003). Axoplasmic importins enable retrograde

injury signaling in lesioned nerve. Neuron, 40, 1095–1104.

253. Hargreaves, K., Dubner, R., Brown, F., Flores, C., & Joris, J. (1988). A new and sensitive

method for measuring thermal nociception in cutaneous hyperalgesia. Pain, 32, 77–88.

254. Haymaker, W. & Woodhall, B. (1953). Peripheral Nerve Injuries, Principles of Diagnosis.

Philadelphia: W. B. Saunders.

255. He, Y. & Baas, P. W. (2003). Growing and working with peripheral neurons. Methods Cell

Biol., 71, 17–35.

256. Hefti, F., Dento, T. L., Knusel, B, & Lapchak, P. A. (1993). Neurotrophic factors: what are

they and what are they doing? In Neurtrophic Factors, ed. S. E., Loughlin & J. H. Fallon.

Toronto: Academic Press, pp. 25–49.

257. Heidenreich, K. A. (1993). Insulin and IGF-I receptor signaling in cultured neurons. Ann.

N.Y. Acad. Sci., 692, 72–88.

234 References

258. Heine, W., Conant, K., Griffin, J. W., & Hoke, A. (2004). Transplanted neural stem

cells promote axonal regeneration through chronically denervated peripheral nerves.

Exp. Neurol., 189, 231–240.

259. Heinrich, P. C., Behrmann, I., Haan, S. et al. (2003). Principles of interleukin (IL)-6-type

cytokine signalling and its regulation. Biochem. J., 374, 1–20.

260. Heinrich, P. C., Behrmann, I., Muller-Newen, G., Schaper, F., & Graeve, L. (1998).

Interleukin-6-type cytokine signalling through the gp130/Jak/STAT pathway. Biochem. J.,

334 (2), 297–314.

261. Helgren, M. E., Squinto, S. P., Davis, H. L. et al. (1994). Trophic effect of ciliary

neurotrophic factor on denervated skeletal muscle. Cell, 76, 493–504.

262. Hendry, I. A., Stockel, K., Thoenen, H., & Iversen, L. L. (1974). The retrograde axonal

transport of nerve growth factor. Brain Res., 68, 103–121.

263. Henrich, M., Hoffmann, K., Konig, P. et al. (2002). Sensory neurons respond to hypoxia

with NO production associated with mitochondria. Mol. Cell Neurosci., 20, 307–322.

264. Herrup, K., Neve, R., Ackerman, S. L., & Copani, A. (2004). Divide and die: cell cycle

events as triggers of nerve cell death. J. Neurosci., 24, 9232–9239.

265. Hess, D. T., Patterson, S. I., Smith, D. S., & Skene, J. H. (1993). Neuronal growth

cone collapse and inhibition of protein fatty acylation by nitric oxide. Nature, 366,

562–565.

266. Heumann, R., Korsching, S., Bandtlow, C., & Thoenen, H. (1987). Changes of nerve

growth factor synthesis in nonneuronal cells in response to sciatic nerve transection.

J. Cell Biol., 104, 1623–1631.

267. Hikawa, N., Horie, H., & Takenaka, T. (1993). Macrophage-enhanced neurite

regeneration of adult dorsal root ganglia neurones in culture. NeuroReport, 5, 41–44.

268. Hildebrand, C., Kocsis, J. D., Berglund, S., & Waxman, S. G. (1985). Myelin sheath

remodelling in regenerated rat sciatic nerve. Brain Res., 358, 163–170.

269. Hildebrand, C., Mustafa, G. Y., Bowe, C., & Kocsis, J. D. (1987). Nodal spacing along

regenerated axons following a crush lesion of the developing rat sciatic nerve. Brain

Res., 429, 147–154.

270. Hiraga, A., Kuwabara, S., Doya, H. et al. (2006). Rho-kinase inhibition enhances axonal

regeneration after peripheral nerve injury. J. Peripher. Nerv. Syst., 11, 217–224.

271. Ho, P. R., Coan, G. M., Cheng, E. T. et al. (1998). Repair with collagen tubules linked with

brain-derived neurotrophic factor and ciliary neurotrophic factor in a rat sciatic nerve

injury model. Arch. Otolaryngol. Head Neck Surg., 124, 761–766.

272. Ho, W. H., Armanini, M. P., Nuijens, A., Phillips, H. S., & Osheroff, P. L. (1995). Sensory

and motor neuron-derived factor. A novel heregulin variant highly expressed in

sensory and motor neurons. J. Biol. Chem., 270, 26722.

273. Hoang, T. X., Pikov, V., & Havton, L. A. (2006). Functional reinnervation of the rat lower

urinary tract after cauda equina injury and repair. J. Neurosci., 26, 8672–8679.

274. Hobson, M. I., Green, C. J., & Terenghi, G. (2000). VEGF enhances intraneural

angiogenesis and improves nerve regeneration after axotomy. J. Anat., 197 (4),

591–605.

275. Hogan, E. L., Dawson, D. M., & Romanul, F. C. A. (1965). Enzymatic changes in

denervated muscle. II Biochemical studies. Arch. Neurol., 13, 274–282.

235References

276. Hohn, A., Leibrock, J., Bailey, K., & Barde, Y. A. (1990). Identification and

characterization of a novel member of the nerve growth factor/brain-derived

neurotrophic factor family. Nature, 344, 339–341.

277. Hoke, A., Cheng, C., & Zochodne, D. W. (2000). Expression of glial cell line-derived

neurotrophic factor family of growth factors in peripheral nerve injury in rats.

NeuroReport, 11, 1651–1654.

278. Hoke, A., Gordon, T., Zochodne, D. W., & Sulaiman, O. A. (2002). A decline in glial

cell-line-derived neurotrophic factor expression is associated with impaired

regeneration after long-term Schwann cell denervation. Exp. Neurol., 173, 77–85.

279. Hoke, A., Ho, T., Crawford, T. O. et al. (2003). Glial cell line-derived neurotrophic factor

alters axon Schwann cell units and promotes myelination in unmyelinated nerve

fibers. J. Neurosci., 23, 561–567.

280. Hoke, A., Redett, R., Hameed, H. et al. (2006). Schwann cells express motor and sensory

phenotypes that regulate axon regeneration. J. Neurosci., 26, 9646–9655.

281. Hoke, A., Sun, H., Gordon, T., & Zochodne, D. W. (2001). Do denervated peripheral

nerve trunks become ischemic? Exp. Neurol., 172, 398–406.

282. Holmes, F. E., Mahoney, S., King, V. R. et al. (2000). Targeted disruption of the galanin

gene reduces the number of sensory neurons and their regenerative capacity. Proc.

Natl. Acad. Sci. USA, 97, 11563–11568.

283. Holmes, M., Maysinger, D., Foerster, A. et al. (2003). Neotrofin, a novel purine that

induces NGF-dependent nociceptive nerve sprouting but not hyperalgesia in adult rat

skin. Mol. Cell Neurosci., 24, 568–580.

284. Honig, M. G., Lance-Jones, C., & Landmesser, L. (1986). The development of sensory

projection patterns in embryonic chick hindlimb under experimental conditions. Dev.

Biol., 118, 532–548.

285. Horie, H., Inagaki, Y., Sohma, Y. et al. (1999). Galectin-1 regulates initial axonal growth

in peripheral nerves after axotomy. J. Neurosci., 19, 9964–9974.

286. Horie, H., Kadoya, T., Hikawa, N. et al. (2004). Oxidized galectin-1 stimulates

macrophages to promote axonal regeneration in peripheral nerves after axotomy.

J. Neurosci., 24, 1873–1880.

287. Horie, H., Sakai, I., Akahori, Y., & Kadoya, T. (1997). IL-1 beta enhances neurite regeneration

from transected-nerve terminals of adult rat DRG. NeuroReport, 8, 1955–1959.

288. Hotta, N., Koh, N., Sakakibara, F. et al. (1996). Effects of beraprost sodium and insulin

on the electroretinogram, nerve conduction, and nerve blood flow in rats with

streptozotocin-induced diabetes. Diabetes, 45, 361–366.

289. Hotta, N., Koh, N., Sakakibara, F. et al. (1996). Effect of propionyl-L-carnitine on motor

nerve conduction, autonomic cardiac function, and nerve blood flow in rats with

streptozotocin-induced diabetes: comparison with an aldose reductase inhibitor.

J. Pharmacol. Exp. Ther., 276, 49–55.

290. Hotta, N., Koh, N., Sakakibara, F. et al. (1995). Prevention of abnormalities in motor

nerve conduction and nerve blood-flow by a prostacyclin analog, beraprost sodium,

in streptozotocin-induced diabetic rats. Prostaglandins, 49, 339–349.

291. Huang, C. S., Zhou, J., Feng, A. K. et al. (1999). Nerve growth factor signaling in

caveolae-like domains at the plasma membrane. J. Biol. Chem., 274, 36707–36714.

236 References

292. Hynes, R. O. (2002). Integrins: bidirectional, allosteric signaling machines. Cell,

110, 673–687.

293. Ibanez, C. F. (2002). Jekyll-Hyde neurotrophins: the story of proNGF. Trends Neurosci.,

25, 284–286.

294. Ibanez, C. F., Ilag, L. L., Murray-Rust, J., & Persson, H. (1993). An extended surface of

binding to Trk tyrosine kinase receptors in NGF and BDNF allows the engineering of

a multifunctional pan-neurotrophin. EMBO J., 12, 2281–2293.

295. Ide, C. (1996). Peripheral nerve regeneration. Neurosci. Res., 25, 101–121.

296. Igarashi, M., Li, W. W., Sudo, Y., & Fishman, M. C. (1995). Ligand-induced growth cone

collapse: amplification and blockade by variant GAP-43 peptides. J. Neurosci., 15,

5660–5667.

297. Ingoglia, N. A. & Murray, M. (2001). Axonal Regeneration in the Central Nervous System.

New York: Marcel Dekker.

298. Inserra, M. M., Yao, M., Murray, R., & Terris, D. J. (2000). Peripheral nerve regeneration

in interleukin 6-deficient mice. Arch. Otolaryngol. Head Neck Surg., 126, 1112–1116.

299. Ip, N. Y. & Yancopoulos, G. D. (1992). Ciliary neurotrophic factor and its receptor

complex. Prog. Growth Factor Res., 4, 139–155.

300. Ishii, D. N. & Lupien, S. B. (1995). Insulin-like growth factors protect against diabetic

neuropathy: effects on sensory nerve regeneration in rats. J. Neurosci. Res., 40, 138–144.

301. Jackson, P. C. & Diamond, J. (1984). Temporal and spatial constraints on the collateral

sprouting of low-threshold mechanosensory nerves in the skin of rats. J. Comp. Neurol.,

226, 336–345.

302. Jain, A., Brady-Kalnay, S. M., & Bellamkonda, R. V. (2004). Modulation of Rho GTPase

activity alleviates chondroitin sulfate proteoglycan-dependent inhibition of neurite

extension. J. Neurosci. Res., 77, 299–307.

303. Jander, S., Lausberg, F., & Stoll, G. (2001). Differential recruitment of CD8þmacrophages during Wallerian degeneration in the peripheral and central nervous

system. Brain Pathol., 11, 27–38.

304. Jander, S., Pohl, J., Gillen, C., & Stoll, G. (1996). Differential expression of

interleukin-1o mRNA in Wallerian degeneration and immune-mediated inflammation

of the rat peripheral nervous system. J. Neurosci. Res., 43, 254–259.

305. Janig, W. & Lisney, S. J. W. (1989). Small diameter myelinated afferents produce

vasodilatation but not plasma extravasation in rat skin. J. Physiol., 415, 477–486.

306. Jaworski, J., Spangler, S., Seeburg, D. P., Hoogenraad, C. C., & Sheng, M. (2005). Control

of dendritic arborization by the phosphoinositide-3’-kinase-Akt-mammalian target of

rapamycin pathway. J. Neurosci., 25, 11300–11312.

307. Jessen, K. R. & Mirsky, R. (1984). Nonmyelin-forming Schwann cells coexpress surface

proteins and intermediate filaments not found in myelin-forming cells: a study of

Ran-2, A5E3 antigen and glial fibrillary acidic protein. J. Neurocytol., 13, 923–934.

308. Jessen, K. R., Morgan, L., Stewart, H. J., & Mirsky, R. (1990). Three markers of adult

non-myelin-forming Schwann cells, 217c(Ran-1), A5E3 and GFAP: development and

regulation by neuron- Schwann cell interactions. Development, 109, 91–103.

309. Ji, R. R., Zhang, Q., Zhang, X. et al. (1995). Prominent expression of bFGF in dorsal root

ganglia after axotomy. Eur. J. Neurosci., 7, 2458–2468.

237References

310. Jiang, Y., McLennan, I. S., Koishi, K., & Hendry, I. A. (2000). Transforming growth

factor-beta 2 is anterogradely and retrogradely transported in motoneurons and

up-regulated after nerve injury. Neuroscience, 97, 735–742.

311. Jin, M., Guan, C. B., Jiang, Y. A. et al. (2005). Ca2þ-dependent regulation of rho GTPases

triggers turning of nerve growth cones. J. Neurosci., 25, 2338–2347.

312. Joffe, M., Savage, N., & Isaacs, H. (1981). Biochemical functioning of mitochondria

in normal and denervated mammalian skeletal muscle. Muscle Nerve, 4, 514–519.

313. Kadomatsu, K. & Muramatsu, T. (2004). Midkine and pleiotrophin in neural

development and cancer. Cancer Lett., 204, 127–143.

314. Kaku, D. A., Malamut, R. I., Frey, D. J., & Parry, G. J. (1993). Conduction block as an early

sign of reversible injury in ischemic monomelic neuropathy. Neurology, 43 (6), 1126.

315. Kalichman, M. W. & Lalonde, A. W. (1991). Experimental nerve ischemia and injury

produced by cocaine and procaine. Brain Res., 565, 34–41.

316. Kamiguchi, H. & Lemmon, V. (2000). Recycling of the cell adhesion molecule L1 in

axonal growth cones. J. Neurosci., 20, 3676–3686.

317. Kamijo, M., Merry, A. C., Akdas, G., Cherian, P. V., & Sima, A. A. (1996). Nerve fiber

regeneration following axotomy in the diabetic biobreeding Worcester rat: the effect

of ARI treatment. J. Diabetes Complications, 10, 183–191.

318. Kaneda, N., Talukder, A. H., Nishiyama, H., Koizumi, S., & Muramatsu, T. (1996).

Midkine, a heparin-binding growth/differentiation factor, exhibits nerve cell adhesion

and guidance activity for neurite outgrowth in vitro. J. Biochem. (Tokyo), 119, 1150–1156.

319. Kang, H., Tian, L., & Thompson, W. (2003). Terminal Schwann cells guide the

reinnervation of muscle after nerve injury. J. Neurocytol., 32, 975–985.

320. Kanje, M., Lundborg, G., & Edstrom, A. (1988). A new method for studies of the effects

of locally applied drugs on peripheral nerve regeneration in vivo. Brain Res., 439,

116–121.

321. Kaplan, D. R. & Miller, F. D. (2000). Neurotrophin signal transduction in the nervous

system. Curr. Opin. Neurobiol., 10, 381–391.

322. Karabay, A., Yu, W., Solowska, J. M., Baird, D. H., & Baas, P. W. (2004). Axonal growth

is sensitive to the levels of katanin, a protein that severs microtubules. J. Neurosci., 24,

5778–5788.

323. Karagiannis, S. N., King, R. H., & Thomas, P. K. (1997). Colocalisation of insulin and

IGF-1 receptors in cultured rat sensory and sympathetic ganglion cells. J. Anat., 191,

431–440.

324. Karchewski, L. A., Gratto, K. A., Wetmore, C., & Verge, V. M. (2002). Dynamic patterns

of BDNF expression in injured sensory neurons: differential modulation by NGF and

NT-3. Eur. J. Neurosci., 16, 1449–1462.

325. Karchewski, L. A., Kim, F. A., Johnston, J., McKnight, R. M., and Verge, V. M. K.

(1999). Anatomical evidence supporting the potential for modulation by multiple

neurotrophins in the majority of adult lumbar sensory neurons. J. Comp. Neurol.,

413, 327–341.

326. Katayama, Y., Montenegro, R., Freier, T. et al. (2006). Coil-reinforced hydrogel tubes

promote nerve regeneration equivalent to that of nerve autografts. Biomaterials,

27, 505–518.

238 References

327. Kato, H., Wanaka, A., & Tohyama, M. (1992). Co-localization of basic fibroblast growth

factor-like immunoreactivity and its receptor mRNA in the rat spinal cord and the

dorsal root ganglion. Brain Res, 576, 351–354.

328. Katsuki, H., & Okuda, S. (1995). Arachidonic acid as a neurotoxic and neurotrophic

substance. Prog. Neurobiol, 46, 607–636.

329. Kawano, S., Okajima, S., Mizoguchi, A. et al. (1997). Immunocytochemical distribution

of Ca2þ-independent protein kinase C subtypes (delta, epsilon, and zeta) in

regenerating axonal growth cones of rat peripheral nerve. Neuroscience, 81, 263–273.

330. Kawasaki, T., Oka, N., Tachibana, H., Akiguchi, I., & Shibasaki, H. (2003). Oct6,

a transcription factor controlling myelination, is a marker for active nerve

regeneration in peripheral neuropathies. Acta Neuropathol. (Berl.), 105, 203–208.

331. Kemp, S. W., Walsh, S. K., Zochodne, D. W., & Midha, R. (2007). A novel method for

establishing daily in vivo concentration gradients of soluble nerve growth factor

(NGF). J. Neurosci. Methods, 165, 83–88.

332. Kennedy, J. M. & Zochodne, D. W. (2000). The regenerative deficit of peripheral nerves

in experimental diabetes: its extent, timing and possible mechanisms. Brain, 123,

2118–2129.

333. Kennedy, J. M. & Zochodne, D. W. (2002). Influence of experimental diabetes on the

microcirculation of injured peripheral nerve. Functional and morphological aspects.

Diabetes, 51, 2233–2240.

334. Kennedy, J. M. & Zochodne, D. W. (2004). Heightened pain and delayed regeneration

of galanin axons in diabetic mice. NeuroReport, 15, 807–810.

335. Kennedy, J. M. & Zochodne, D. W. (2005). Experimental diabetic neuropathy and

spontaneous recovery: is there irreparable damage? Diabetes, 54, 830–837.

336. Kennedy, J. M. & Zochodne, D. W. (2005). Impaired peripheral nerve regeneration in

diabetes mellitus. J. Peripher. Nerv. Syst., 10, 144–157.

337. Kerezoudi, E. & Thomas, P. K. (1999). Influence of age on regeneration in the peripheral

nervous system. Gerontology, 45, 301–306.

338. Keswani, S. C., Buldanlioglu,U., Fischer, A. et al. (2004). Anovel endogenous erythropoietin

mediated pathway prevents axonal degeneration. Ann. Neurol., 56, 815–826.

339. Kilmer, S. L. & Carlsen, R. C. (1984). Forskolin activation of adenylate cyclase in vivo

stimulates nerve regeneration. Nature, 307, 455–457.

340. Kilmer, S. L. & Carlsen, R. C. (1987). Chronic infusion of agents that increase cyclic AMP

concentration enhances the regeneration of mammalian peripheral nerves in vivo.

Exp. Neurol., 95, 357–367.

341. Kim, B., Cheng, H. L., Margolis, B., & Feldman, E. L. (1998). Insulin receptor substrate 2

and Shc play different roles in insulin-like growth factor I signaling. J. Biol. Chem., 273,

34543–34550.

342. Kim, K. J., Yoon, Y. W., & Chung, J. M. (1997). Comparison of three rodent neuropathic

pain models. Exp. Brain Res., 113, 200–206.

343. Kim, Y. S., Furman, S., Sink, H., & VanBerkum, M. F. (2001). Calmodulin and profilin

coregulate axon outgrowth in Drosophila. J. Neurobiol., 47, 26–38.

344. Kimpinski, K., Campenot, R. B., & Mearow, K. (1997). Effects of the neurotrophins nerve

growth factor, neurotrophin-3, and brain-derived neurotrophic factor (BDNF) on

239References

neurite growth from adult sensory neurons in compartmented cultures. J. Neurobiol.,

33, 395–410.

345. Kimura, J. (2001). Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice,

3rd edn. New York: Oxford University Press.

346. King, R. H. M. (1999). Atlas of Peripheral Nerve Pathology. London: Arnold.

347. Kishino, A., Ishige, Y., Tatsuno, T., Nakayama, C., & Noguchi, H. (1997). BDNF prevents

and reverses adult rat motor neuron degeneration and induces axonal outgrowth.

Exp. Neurol., 144, 273–286.

348. Klein, H. W., Kilmer, S., & Carlsen, R. C. (1989). Enhancement of peripheral nerve

regeneration by pharmacological activation of the cyclic AMP second messenger

system. Microsurgery, 10, 122–125.

349. Kline, D. G. & Hudson, A. (1995). Nerve Injuries: Operative Results for Major Nerve Injuries,

Entrapments, and Tumors. Toronto: W. B. Saunders.

350. Kobayashi, N. R., Fan, D. P., Giehl, K. M. et al. (1997). BDNF and NT-4/5 prevent

atrophy of rat rubrospinal neurons after cervical axotomy, stimulate GAP-43 and

Talpha1-tubulin mRNA expression, and promote axonal regeneration. J. Neurosci.,

17, 9583–9595.

351. Kobbert, C., Apps, R., Bechmann, I. et al. (2000). Current concepts in neuroanatomical

tracing. Prog. Neurobiol., 62, 327–351.

352. Koenig, E., Martin, R., Titmus, M., & Sotelo-Silveira, J. R. (2000). Cryptic peripheral

ribosomal domains distributed intermittently along mammalian myelinated axons.

J. Neurosci., 20, 8390–8400.

353. Kogawa, S., Yasuda, H., Terada, M., Maeda, K., & Kikkawa, R. (2000). Apoptosis and

impaired axonal regeneration of sensory neurons after nerve crush in diabetic rats.


354. Kolston, J. & Lisney, S. J. W. (1993). A study of vasodilator responses evoked by

antidromic stimulation of A delta afferent nerve fibers supplying normal and

reinnervated rat skin. Microvasc. Res., 46, 143–157.

355. Korthals, J., Gieron, M., & Wisniewski, H. (1989). Nerve regeneration patterns after

acute ischemic injury. Neurology, 39, 932.

356. Korthals, J. K., Korthals, M. A., & Wisniewski, H. M. (1978). Peripheral nerve ischemia:

Part 2. Accumulation of organelles. Ann. Neurol., 4, 487–498.

357. Korthals, J. K., Maki, T., & Gieron, M. A. (1985). Nerve and muscle vulnerability to

ischemia. J. Neurol. Sci., 71, 283–290.

358. Korthals, J. K., Maki, T., Korthals, M. A., & Prockop, L. D. (1996). Nerve and muscle

damage after experimental thrombosis of large artery. Electrophysiology and

morphology. J. Neurol. Sci., 136, 24–30.

359. Korthals, J. K. & Wisniewski, H. M. (1975). Peripheral nerve ischemia. Part 1.

Experimental model. J. Neurol. Sci., 24, 65–76.

360. Kraft, G. H. (1996). Are fibrillation potentials and positive sharp waves the same?

No. Muscle Nerve, 19 (2), 216–220.

361. Krasnoselsky, A., Massay, M. J., DeFrances, M. C. et al. (1994). Hepatocyte growth factor

is a mitogen for Schwann cells and is present in neurofibromas. J. Neurosci., 14,

7284–7290.

240 References

362. Krekoski, C. A., Neubauer, D., Zuo, J., & Muir, D. (2001). Axonal regeneration into

acellular nerve grafts is enhanced by degradation of chondroitin sulfate proteoglycan.

J. Neurosci., 21, 6206–6213.

363. Kreulen, D. L. (2005). Neurobiology of autonomic ganglia. In Peripheral Neuropathy,

4th edn., ed. P. J. Dyck & P. K. Thomas. Philadelphia: Elsevier Saunders, pp. 233–248.

364. Kumagai, K., Ushiki, T., Tohyama, K., Arakawa, M., & Ide, C. (1990). Regenerating axons

and their growth cones observed by scanning electron microscopy. J. Electron Microsc.

(Tokyo), 39, 108–114.

365. Kuo, L. T., Simpson, A., Schanzer, A. et al. (2005). Effects of systemically administered

NT-3 on sensory neuron loss and nestin expression following axotomy. J. Comp. Neurol.,

482, 320–332.

366. Kuwako, K. I., Hosokawa, A., Nishimura, I. et al. (2005). Disruption of the paternal

necdin gene diminishes TrkA signaling for sensory neuron survival. J. Neurosci., 25,

7090–7099.

367. Laferriere, N. B., MacRae, T. H., & Brown, D. L. (1997). Tubulin synthesis and assembly

in differentiating neurons. Biochem. Cell Biol., 75, 103–117.

368. Lagerlund, T. D. & Low, P. A. (1987). A mathematical simulation of oxygen delivery in

rat peripheral nerve. Microvasc. Res., 34, 211–222.

369. Lagerlund, T. D. & Low, P. A. (1994). Mathematical modeling of hydrogen clearance

blood flow measurements in peripheral nerve. Comput. Biol. Med., 24, 77–89.

370. Lago, N. & Navarro, X. (2007). Evaluation of the long-term regenerative potential in an

experimental nerve amputee model. J. Peripher. Nerv. Syst., 12, 108–120.

371. Lai, K. O., Fu, W. Y., Ip, F. C. F., & Ip, N. Y. (1998). Cloning and expression of a novel

neurotrophin, NT-7, from carp. Mol. Cell Neurosci., 11, 64–76.

372. Landreth, G. E. (2006). Growth factors. In Basic Neurochemistry, 7th edn., ed. G. J. Siegel,

R. W. Albers, S. T. Brady, & D. L. Price. New York: Academic Press, pp. 471–484.

373. Langton, B. S. (2002). A First Step: Understanding Guillain–Barre Syndrome. Trafford

Publishing.

374. Langton, B. S., Ondrich, S., & Hill, P. (2006). Guillain–Barre Syndrome. 5 years later.

Trafford Publishing.

375. Lariviere, R. C. & Julien, J. P. (2004). Functions of intermediate filaments in neuronal

development and disease. J. Neurobiol., 58, 131–148.

376. Latchman, D. S. (2005). HSP27 and cell survival in neurones. Int. J. Hyperthermia, 21,

393–402.

377. Lauria, G., Cornblath, D. R., Johansson, O. et al. (2005). EFNS guidelines on the use of

skin biopsy in the diagnosis of peripheral neuropathy. Eur. J. Neurol., 12, 747–758.

378. Laux, T., Fukami, K., Thelen, M. et al. (2000). GAP43, MARCKS, and CAP23 modulate

PI(4,5)P(2) at plasmalemmal rafts, and regulate cell cortex actin dynamics through

a common mechanism. J. Cell Biol., 149, 1455–1472.

379. Lawson, S. N. (2005). The peipheral sensory nervous system: dorsal root ganglion

neurons. In Peripheral Neuropathy, 4th edn., ed. P. J. Dyck & P. K. Thomas. Philadelphia:

Elsevier Saunders, pp. 163–202.

380. LeBlanc, A. C. & Poduslo, J. F. (1990). Axonal modulation of myelin gene expression

in the peripheral nerve. J. Neurosci. Res., 26, 317–326.

241References

381. Leclere, P. G., Norman, E., Groutsi, F. et al. (2007). Impaired axonal regeneration by

isolectin B4-binding dorsal root ganglion neurons in vitro. J. Neurosci., 27, 1190–1199.

382. Lee, A. C., Yu, V. M., Lowe, J. B. et al. (2003). Controlled release of nerve growth factor

enhances sciatic nerve regeneration. Exp. Neurol., 184, 295–303.

383. Lee, R., Kermani, P., Teng, K. K., & Hempstead, B. L. (2001). Regulation of cell survival by

secreted proneurotrophins. Science, 294, 1945–1948.

384. Lee, Y. S., Baratta, J., Yu, J., Lin, V. W., & Robertson, R. T. (2002). AFGF promotes axonal

growth in rat spinal cord organotypic slice co-cultures. J. Neurotrauma., 19, 357–367.

385. Lehmann, H. C., Lopez, P. H., Zhang, G. et al. (2007). Passive immunization with

anti-ganglioside antibodies directly inhibits axon regeneration in an animal model.

J. Neurosci., 27, 27–34.

386. Leibrock, J., Lottspeich, F., Hohn, A. et al. (1989). Molecular cloning and expression of

brain-derived neurotrophic factor. Nature, 341, 149–152.

387. Levavasseur, F., Zhu, Q., & Julien, J. P. (1999). No requirement of alpha-internexin for

nervous system development and for radial growth of axons. Brain Res. Mol. Brain Res.,

69, 104–112.

388. Levi-Montalcini, R. (1987). The nerve growth factor 35 years later. Science, 237, 1154–1162.

389. Levi-Montalcini, R. & Hamburger, V. (1951). Selective growth stimulating effects of

mouse sarcoma on the sensory and sympathetic nervous system of the chick embryo.

J. Exp. Zool., 116, 321–362.

390. Levi-Montalcini, R. & Hamburger, V. (1953). A diffusible agent of mouse sarcoma,

producing hyperplasia of sympathetic ganglia and hyperneurotization of viscera in

the chick embro. J. Exp. Zool., 123, 233–287.

391. Levy, D., Hoke, A., & Zochodne, D. W. (1999). Local expression of inducible nitric oxide

synthase in an animal model of neuropathic pain. Neurosci. Lett., 260, 207–209.

392. Levy, D., Kubes, P., & Zochodne, D. W. (2001). Delayed peripheral nerve degeneration,

regeneration, and pain in mice lacking inducible nitric oxide synthase. J. Neuropathol.

Exp. Neurol., 60, 411–421.

393. Levy, D., Tal, M., Hoke, A., & Zochodne, D. W. (2000). Transient action of the endothelial

constitutive nitric oxide synthase (ecNOS) mediates the development of thermal

hypersensitivity following peripheral nerve injury. Eur. J. Neurosci., 12, 2323–2332.

394. Levy, D. & Zochodne, D. W. (1998). Local nitric oxide synthase activity in a model of

neuropathic pain. Eur. J. Neurosci., 10, 1846–1855.

395. Lewis, T., Harris, K. E., & Grant, R. T. (1927). Observations relating to the influence

of the cutaneous nerves on various reactions of the cutaneous vessels. Heart, 14, 1–17.

396. Li, G. D., Wo, Y., Zhong, M. F. et al. (2002). Expression of fibroblast growth factors in rat

dorsal root ganglion neurons and regulation after peripheral nerve injury.


397. Li, H., Terenghi, G., & Hall, S. M. (1997). Effects of delayed re-innervation on the

expression of c-ergB receptors by chronically denervated rat Schwann cells in vivo.

Glia, 20, 333–347.

398. Li, L., Oppenheim, R. W., Lei, M., & Houenou, L. J. (1994). Neurotrophic agents prevent

motoneuron death following sciatic nerve section in the neonatal mouse. J. Neurobiol.,

25, 759–766.

242 References

399. Li, R. (2005). Neuronal polarity: until GSK-3 do us part. Curr. Biol., 15, R198–R200.

400. Li, X.-G. & Zochodne, D. W. (2003). Microvacuolar neuronopathy is a post-mortem

artifact of sensory neurons. J. Neurocytol., 32, 393–398.

401. Li, X.-Q., Verge, V. M. K., Johnston, J. M., & Zochodne, D. W. (2004). CGRP peptide and

regenerating sensory axons. J. Neuropathol. Exp. Neurol., 63, 1092–1103.

402. Li, Y. S., Milner, P. G., Chauhan, A. K. et al. (1990). Cloning and expression of a

developmentally regulated protein that induces mitogenic and neurite outgrowth

activity. Science, 250, 1690–1694.

403. Lieberman, A. R. (1971). The axon reaction: a review of the principal features of

perikaryal responses to axon injury. Int. Rev. Neurobiol., 14, 49–124.

404. Liefner, M., Siebert, H., Sachse, T. et al. (2000). The role of TNF-alpha during Wallerian

degeneration. J. Neuroimmunol., 108, 147–152.

405. Lin, H., Bao, J., Sung, Y. J., Walters, E. T., & Ambron, R. T. (2003). Rapid electrical and

delayed molecular signals regulate the serum response element after nerve injury:

convergence of injury and learning signals. J. Neurobiol., 57, 204–220.

406. Lin, L. F., Doherty, D. H., Lile, J. D., Bektesh, S., & Collins, F. (1993). GDNF: a glial cell

line-derived neurotrophic factor for midbrain dopaminergic neurons. Science, 260,

1130–1132.

407. Lin, L. F., Zhang, T. J., Collins, F., & Armes, L. G. (1994). Purification and initial

characterization of rat B49 glial cell line-derived neurotrophic factor. J. Neurochem.,

63, 758–768.

408. Lindholm, D., Heumann, R., Meyer, M., & Thoenen, H. (1987). Interleukin-1 regulates

synthesis of nerve growth factor in non-neuronal cells of rat sciatic nerve. Nature, 330,

658–659.

409. Lino, M. M., Atanasoski, S., Kvajo, M. et al. (2007). Mice lacking protease nexin-1 show

delayed structural and functional recovery after sciatic nerve crush. J. Neurosci., 27,

3677–3685.

410. Liu, C. N., Amir, R., & Devor, M. (1999). Effect of age and nerve injury on cross-

excitation among sensory neurons in rat dorsal root ganglia. Neurosci. Lett., 259, 95–98.

411. Liu, R. Y., Schmid, R. S., Snider, W. D., & Maness, P. F. (2002). NGF enhances sensory

axon growth induced by laminin but not by the L1 cell adhesion molecule. Mol. Cell

Neurosci., 20, 2–12.

412. Lohof, A. M., Quillan, M., Dan, Y., & Poo, M.-M. (1992). Asymmetric modulation of

cytosolic cAMP activity induces growth cone turning. J. Neurosci., 12, 1253–1261.

413. Longo, F. M., Powell, H. C., Lebeau, J. et al. (1986). Delayed nerve regeneration in

streptozotocin diabetic rats. Muscle Nerve, 9, 385–393.

414. Love, A., Cotter, M. A., & Cameron, N. E. (1995). Impaired myelinated fiber regeneration

following freeze-injury in rats with streptozotocin-induced diabetes: involvement of

the polyol pathway. Brain Res., 703, 105–110.

415. Low, P. A., Lagerlund, T. D., & McManis, P. G. (1989). Nerve blood flow and oxygen

delivery in normal, diabetic, and ischemic neuropathy. Int. Rev. Neurobiol., 31, 355–438.

416. Low, P. A. & Tuck, R. R. (1984). Effects of changes of blood pressure, respiratory

acidosis and hypoxia on blood flow in the sciatic nerve of the rat. J. Physiol., 347,

513–524.

243References

417. Lu, X. & Richardson, P. M. (1991). Inflammation near the nerve cell body enhances

axonal regeneration. J. Neurosci., 11, 972–978.

418. Lubinska, L. & Olekiewicz, M. (1959). The rate of regeneration of amphibian peripheral

nerves at different temperatures. Acta Biol. Exp. (Warsaw), 15, 125–145.

419. Lumpkin, E. A. & Caterina, M. J. (2007). Mechanisms of sensory transduction in the

skin. Nature, 445, 858–865.

420. Lundborg, G. (1975). Structure and function of the intraneural microvessels as related

to trauma, edema formation, and nerve function. J. Bone Joint Surg. AM., 57, 938–948.

421. Lundborg, G. (2004). Nerve Injury and Repair. Regeneration, Reconstruction and Cortical

Remodeling, 2nd edn. Elsevier.

422. Lundborg, G., Dahlin, L., Dohi, D., Kanje, M., & Terada, N. (1997). A new type of

“bioartificial” nerve graft for bridging extended defects in nerves. J. Hand Surg. [Br].,

22, 299–303.

423. Lundborg, G., Dahlin, L. B., & Danielsen, N. (1991). Ulnar nerve repair by the silicone

chamber technique. Case report. Scand. J. Plast. Reconstr. Surg., 25, 79–82.

424. Lundborg, G., Dahlin, L. B., Danielsen, N. et al. (1982). Nerve regeneration in silicone

chambers: influence of gap length and of distal stump components. Exp. Neurol., 76,

361–375.

425. Lundborg, G., Dahlin, L. B., Danielsen, N. et al. (1982). Nerve regeneration across an

extended gap: a neurobiological view of nerve repair and the possible involvement

of neuronotrophic factors. J. Hand Surg. [Am]., 7, 580–587.

426. Lundborg, G., Dahlin, L. B., Danielsen, N. P., Hansson, H. A., & Larsson, K. (1981).

Reorganization and orientation of regenerating nerve fibres, perineurium, and

epineurium in preformed mesothelial tubes – an experimental study on the sciatic

nerve of rats. J. Neurosci. Res., 6, 265–281.

427. Lundborg, G., Gelberman, R. H., Longo, F. M., Powell, H. C., & Varon, S. (1982). In vivo

regenerationofcutnervesencased insiliconetubes. J.Neuropathol. Exp.Neurol.,41, 412–422.

428. Lundborg, G. & Hansson, H. A. (1979). Regeneration of peripheral nerve through

a preformed tissue space. Preliminary observations on the reorganization of

regenerating nerve fibres and perineurium. Brain Res., 178, 573–576.

429. Luo, L., Jan, L. Y., & Jan, Y. N. (1997). Rho family GTP-binding proteins in growth cone

signalling. Curr. Opin. Neurobiol., 7, 81–86.

430. Luo, L. & O’Leary, D. D. (2005). Axon retraction and degeneration in development and

disease. Ann. Rev. Neurosci., 28, 127–156.

431. Lyons, W. R. & Woodhall, B. (1949). Atlas of Peripheral Nerve Injuries. Philadelphia:

W. B. Saunders.

432. Macica, C. M., Liang, G., Lankford, K. L., & Broadus, A. E. (2006). Induction of

parathyroid hormone-related peptide following peripheral nerve injury: role as a

modulator of Schwann cell phenotype. Glia, 53, 637–648.

433. Mack, T. G., Reiner, M., Beirowski, B. et al. (2001). Wallerian degeneration of injured

axons and synapses is delayed by a Ube4b/Nmnat chimeric gene. Nat. Neurosci., 4,

1199–1206.

434. Maeda, K., Fernyhough, P., & Tomlinson, D. R. (1996). Regenerating sensory neurones

of diabetic rats express reduced levels of mRNA for GAP-43, gamma-preprotachykinin

244 References

and the nerve growth factor receptors, trkA and p75NGFR. Brain Res. Mol. Brain Res.,

37, 166–174.

435. Maina, F. & Klein, R. (1999). Hepatocyte growth factor, a versatile signal for developing

neurons. Nat. Neurosci., 2, 213–217.

436. Maisonpierre, P. C., Belluscio, L., Squinto, S. et al. (1990). Neurotrophin-3: a neurotrophic

factor related to NGF and BDNF. Science, 247, 1446–1451.

437. Malik, R. A., Newrick, P. G., Sharma, A. K. et al. (1989). Microangiopathy in human

diabetic neuropathy: relationship between capillary abnormalities and the severity

of neuropathy. Diabetologia, 32, 92–102.

438. Martinez, J. A., Cunningham, F. A., & Zochodne, D. W. (2005). Appraising mouse nerve:

dessication artifact. J. Peripher. Nerv. Syst., 10, 213–214.

439. Martini, R. (1994). Expression and functional roles of neural cell surface molecules and

extracellular matrix components during development and regeneration of peripheral

nerves. J. Neurocytol., 23, 1–28.

440. Martini, R. & Schachner, M. (1988). Immunoelectron microscopic localization of neural

cell adhesion molecules (L1, N-CAM, and myelin-associated glycoprotein) in

regenerating adult mouse sciatic nerve. J. Cell Biol., 106, 1735–1746.

441. Martini, R., Xin, Y., Schmitz, B., & Schachner, M. (1992). The L2/HNK-1 carbohydrate

epitope is involved in the preferential outgrowth of motor neurons on ventral roots

and motor nerves. Eur. J. Neurosci., 4, 628–639.

442. Mason, C. A. & Wang, L. C. (1997). Growth cone form is behavior-specific and,

consequently, position-specific along the retinal axon pathway. J. Neurosci., 17,

1086–1100.

443. Mason, M. R., Lieberman, A. R., Grenningloh, G., & Anderson, P. N. (2002).

Transcriptional upregulation of SCG10 and CAP-23 is correlated with regeneration

of the axons of peripheral and central neurons in vivo. Mol. Cell Neurosci., 20, 595–615.

444. Massa, S. M., Xie, Y., Yang, T. et al. (2006). Small, nonpeptide p75NTR ligands induce

survival signaling and inhibit proNGF-induced death. J. Neurosci., 26, 5288–5300.

445. Matsumoto, K., Sawa, H., Sato, M. et al. (2002). Distribution of extracellular matrix

tenascin-X in sciatic nerves. Acta Neuropathol. (Berl.), 104, 448–454.

446. Matsunaga, E., Nakamura, H., & Chedotal, A. (2006). Repulsive guidance molecule plays

multiple roles in neuronal differentiation and axon guidance. J. Neurosci., 26, 6082–6088.

447. Matsuoka, I., Nakane, A., & Kurihara, K. (1997). Induction of LIF-mRNA by TGF-beta 1

in Schwann cells. Brain Res., 776, 170–180.

448. Maxfield, E. K., Love, A., Cotter, M. A., & Cameron, N. E. (1995). Nerve function

and regeneration in diabetic rats: effects of ZD-7155, an AT1 receptor antagonist. Am. J.

Physiol., 269, E530–E537.

449. McComas, A. J., Fawcett, P. R. W., Campbell, M. J., & Sica, R. E. P. (1971).

Electrophysiological estimation of the number of motor units within a human

muscle. J. Neurol. Neurosurg. Psychiatry, 34, 121–131.

450. McDonald, D., Cheng, C., Chen, Y. Y., & Zochodne, D. W. (2006). Early events of

peripheral nerve regeneration. Neuron Glia Biol., 2, 139–147.

451. McDonald, D. M. (2007). Supportive and inhibitory influences on the outcome of

peripheral nerve regeneration. Ph.D. thesis, University of Calgary.

245References

452. McDonald, D. S. & Zochodne, D. W. (2003). An injectable nerve regeneration chamber

for studies of unstable soluble growth factors. J. Neurosci. Methods, 122, 171–178.

453. McDonald, N. Q., Lapatto, R., Murray-Rust, J. et al. (1991). New protein fold revealed by

a 2.3-A resolution crystal structure of nerve growth factor. Nature, 354, 411–414.

454. McGraw, T. S., Mickle, J. P., Shaw, G., & Streit, W. J. (2002). Axonally transported

peripheral signals regulate alpha-internexin expression in regenerating motoneurons.

J. Neurosci., 22, 4955–4963.

455. McKenzie, I. A., Biernaskie, J., Toma, J. G., Midha, R., & Miller, F. D. (2006). Skin-derived

precursors generate myelinating Schwann cells for the injured and dysmyelinated

nervous system. J. Neurosci., 26, 6651–6660.

456. McLean, W. G., Pekiner, C., Cullum, N. A., & Casson, I. F. (1992). Posttransiational

modifications of nerve cytoskeletal proteins in experimental diabetes. Mol. Neurobiol.,

6, 225–237.

457. McMahon, S. B., Armanini, M. P., Ling, L. H., & Phillips, H. S. (1994). Expression and

coexpression of Trk receptors in subpopulations of adult primary sensory neurons

projecting to identified peripheral targets. Neuron, 12, 1161–1171.

458. McManis, P. G. & Low, P. A. (1988). Factors affecting the relative viability of

centrifascicular and subperineurial axons in acute peripheral nerve ischemia. Exp.

Neurol., 99, 84–95.

459. McManis, P. G., Schmelzer, J. D., Zollman, P. J., & Low, P. A. (1997). Blood flow and

autoregulation in somatic and autonomic ganglia. Comparison with sciatic

nerve. Brain, 120 (3), 445–449.

460. McQuarrie, I. G. & Grafstein, B. (1973). Axon outgrowth enhanced by a previous nerve

injury. Arch. Neurol., 29, 53–55.

461. McQuarrie, I. G., Grafstein, B., & Gershon, M. D. (1977). Axonal regeneration in

the rat sciatic nerve: effect of a conditioning lesion and of dbcAMP. Brain Res., 132,

443–453.

462. Meier, C., Parmantier, E., Brennan, A., Mirsky, R., & Jessen, K. R. (1999). Developing

Schwann cells acquire the ability to survive without axons by establishing an

autocrine circuit involving insulin-like growth factor, neurotrophin-3, and platelet-

derived growth factor- BB. J. Neurosci., 19, 3847–3859.

463. Metz, G. A. & Whishaw, I. Q. (2002). Cortical and subcortical lesions impair skilled

walking in the ladder rung walking test: a new task to evaluate fore- and hindlimb

stepping, placing, and co-ordination. J. Neurosci. Methods, 115, 169–179.

464. Meyer, D. & Birchmeier, C. (1995). Multiple essential functions of neuregulin in

development. Nature, 378, 386–390.

465. Meyer, G. & Feldman, E. L. (2002). Signaling mechanisms that regulate actin-based

motility processes in the nervous system. J. Neurochem., 83, 490–503.

466. Meyer, M., Matsuoka, I., Wetmore, C., Olson, L., & Thoenen, H. (1992). Enhanced

synthesis of brain-derived neurotrophic factor in the lesioned peripheral nerve:

different mechanisms are responsible for the regulation of BDNF and NGF mRNA.

J. Cell Biol., 119, 45–54.

467. Mi, R., Chen, W., & Hoke, A. (2007). Pleiotrophin is a neurotrophic factor for spinal

motor neurons. Proc. Natl. Acad. Sci. USA, 104, 4664–4669.

246 References

468. Mi, S., Lee, X., Shao, Z. et al. (2004). LINGO-1 is a component of the Nogo-66 receptor/p75

signaling complex. Nat. Neurosci., 7, 221–228.

469. Miao, T., Wu, D., Zhang, Y. et al. (2006). Suppressor of cytokine signaling-3 suppresses

the ability of activated signal transducer and activator of transcription-3 to stimulate

neurite growth in rat primary sensory neurons 1190. J. Neurosci., 26, 9512–9519.

470. Michailov, G. V., Sereda, M. W., Brinkmann, B. G. et al. (2004). Axonal neuregulin-1

regulates myelin sheath thickness. Science, 304, 700–703.

471. Midrio, M. (2006). The denervated muscle: facts and hypotheses. A historical review.

Eur. J. Appl. Physiol., 98, 1–21.

472. Miller, F. D., Tetzlaff, W., Bisby, M. A., Fawcett, J. W., & Milner, R. J. (1989). Rapid

induction of the major embryonic a-Tubulin mRNA, Ta1, during nerve regeneration

in adult rats. J. Neurol. Sci., 9, 1452–1463.

473. Ming, G., Song, H., Berninger, B. et al. (1999). Phospholipase C-gamma and

phosphoinositide 3-kinase mediate cytoplasmic signaling in nerve growth cone

guidance. Neuron, 23, 139–148.

474. Ming, G. L., Song, H. J., Berninger, B. et al. (1997). cAMP-dependent growth cone

guidance by netrin-1. Neuron, 19, 1225–1235.

475. Mirsky, R. & Jessen, K. R. (1999). The neurobiology of Schwann cells. Brain Pathol.,

9, 293–311.

476. Moir, M. S., Wang, M. Z., To, M., Lum, J., & Terris, D. J. (2000). Delayed repair of

transected nerves: effect of brain-derived neurotrophic factor. Arch. Otolaryngol. Head

Neck Surg., 126, 501–505.

477. Morris, J. H., Hudson, A. R., & Weddell, G. (1972). A study of degeneration and

regeneration in the divided rat sciatic nerve based on electron microscopy. I.

The traumatic degeneration of myelin in the proximal stump of the divided nerve.

Z. Zellforsch. Mikrosk. Anat., 124, 76–102.


regeneration in the divided rat sciatic nerve based on electron microscopy. II.

The development of the “regenerating unit”. Z. Zellforsch. Mikrosk. Anat., 124, 103–130.


regeneration in the divided rat sciatic nerve based on electron microscopy. IV.

Changes in fascicular microtopography, perineurium and endoneurial fibroblasts.

Z. Zellforsch. Mikrosk. Anat., 124, 165–203.

480. Morris, J. K., Lin, W., Hauser, C. et al. (1999). Rescue of the cardiac defect in ErbB2

mutant mice reveals essential roles of ErbB2 in peripheral nervous system

development. Neuron, 23, 273–283.

481. Morton, C. R., Lacey, G. R., & Newcombe, R. L. G. (2000). Mechanisms of pain

arising from spinal nerve root compression. In Proceedings of the 9th World Congress

of Pain, ed. M. Devor, M. C. Rowbotham, & Z. Wiesenfeld-Hallin. Seattle: IASP Press,

pp. 689–695.

482. Mosahebi, A., Woodward, B., Wiberg, M., Martin, R., & Terenghi, G. (2001). Retroviral

labeling of Schwann cells: in vitro characterization and in vivo transplantation to

improve peripheral nerve regeneration. Glia, 34, 8–17.

483. Munger, B. L. (1988). The reinnervation of denervated skin. Prog. Brain Res., 74, 259–262.

247References

484. Murakami, K., Namikawa, K., Shimizu, T. et al. (2006). Nerve injury induces the

expression of EXT2, a glycosyltransferase required for heparan sulfate synthesis.


485. Murashov, A. K., Ul, H., I, Hill, C. et al. (2001). Crosstalk between p38, Hsp25 and Akt in

spinal motor neurons after sciatic nerve injury. Brain Res. Mol. Brain Res., 93, 199–208.

486. Murphy, P., Topilko, P., Schneider-Maunoury, S. et al. (1996). The regulation of Krox-20

expression reveals important steps in the control of peripheral glial cell development.

Development, 122, 2847–2857.

487. Murphy, P. G., Borthwick, L. A., Altares, M. et al. (2000). Reciprocal actions of

interleukin-6 and brain-derived neurotrophic factor on rat and mouse primary

sensory neurons. Eur. J. Neurosci., 12, 1891–1899.

488. Murphy, P. G., Borthwick, L. S., Johnston, R. S., Kuchel, G., & Richardson, P. M. (1999).

Nature of the retrograde signal from injured nerves that induces interleukin-6 mRNA

in neurons. J. Neurosci., 19, 3791–3800.

489. Murphy, P. G., Grondin, J., Altares, M., & Richardson, P. M. (1995). Induction of

interleukin-6 in axotomized sensory neurons. J. Neurosci., 15, 5130–5138.

490. Myers, R. R., Heckman, H. M., & Powell, H. C. (1983). Endoneurial fluid is hypertonic.

Results of microanalysis and its significance in neuropathy. J. Neuropathol. Exp. Neurol.,

42, 217–224.

491. Nachemson, A. K., Hansson, H. A., Dahl, D., & Lundborg, G. (1986). Visualization of

regenerating sciatic nerve fibres by neurofilament immunohistochemistry. Acta

Physiol. Scand., 126, 485–489.

492. Nachemson, A. K., Hansson, H. A., & Lundborg, G. (1988). Neurotropism in nerve

regeneration: an immunohistochemical study. Acta Physiol. Scand., 133, 139–148.

493. Nachmias, V. T., Golla, R., Casella, J. F., & Barron-Casella, E. (1996). Cap Z, a calcium

insensitive capping protein in resting and activated platelets. FEBS Lett., 378, 258–262.

494. Nachmias, V. T. & Padykula, H. A. (1958). A histochemical study of normal and

denervated red and white muscles of the rat. J. Biophysic & Biochem. Cytol., 4, 47–59.

495. Nakagawa, T., Miyamoto, O., Janjua, N. A. et al. (2004). Localization of nestin in

amygdaloid kindled rat: an immunoelectron microscopic study. Can. J. Neurol. Sci.,

31, 514–519.

496. Namikawa, K., Honma, M., Abe, K. et al. (2000). Akt/protein kinase B prevents injury-

induced motoneuron death and accelerates axonal regeneration. J. Neurosci., 20,

2875–2886.

497. Namikawa, K., Okamoto, T., Suzuki, A., Konishi, H., & Kiyama, H. (2006). Pancreatitis-

associated protein-III is a novel macrophage chemoattractant implicated in nerve

regeneration. J. Neurosci., 26, 7460–7467.

498. Nandedkar, S. D., Barkhaus, P. E., Sanders, D. B., & Stalberg, E. V. (2000). Some

observations on fibrillations and positive sharp waves. Muscle Nerve, 23, 888–894.

499. Navarro, X. & Kennedy, W. R. (1988). Effect of age on collateral reinnervation of sweat

glands in the mouse. Brain Res., 463, 174–181.

500. Navarro, X., Udina, E., Ceballos, D., & Gold, B. G. (2001). Effects of FK506 on nerve

regeneration and reinnervation after graft or tube repair of long nerve gaps. Muscle

Nerve, 24, 905–915.

248 References

501. Navarro, X., Verdu, E., Wendelschafer-Crabb, G., & Kennedy, W. R. (1997).

Immunohistochemical study of skin reinnervation by regenerative axons. J. Comp.

Neurol., 380, 164–174.

502. Neary, J. T., Rathbone, M. P., Cattabeni, F., Abbracchio, M. P., & Burnstock, G. (1996).

Trophic actions of extracellular nucleotides and nucleosides on glial and neuronal

cells. Trends Neurosci., 19, 13–18.

503. Neumann, S., Bradke, F., Tessier-Lavigne, M., & Basbaum, A. I. (2002). Regeneration of

sensory axons within the injured spinal cord induced by intraganglionic cAMP

elevation. Neuron, 34, 885–893.

504. Neumann, S., Skinner, K., & Basbaum, A. I. (2005). Sustaining intrinsic growth capacity

of adult neurons promotes spinal cord regeneration. Proc. Natl. Acad. Sci. USA, 102,

16848–16852.

505. Ng, J. & Luo, L. (2004). Rho GTPases regulate axon growth through convergent and

divergent signaling pathways. Neuron, 44, 779–793.

506. Niederost, B., Oertle, T., Fritsche, J., McKinney, R. A., & Bandtlow, C. E. (2002). Nogo-A

andmyelin-associated glycoprotein mediate neurite growth inhibition by antagonistic

regulation of RhoA and Rac1. J. Neurosci., 22, 10368–10376.

507. Niwa, R., Nagata-Ohashi, K., Takeichi, M., Mizuno, K., & Uemura, T. (2002). Control of

actin reorganization by Slingshot, a family of phosphatases that dephosphorylate ADF/

cofilin. Cell, 108, 233–246.

508. Nixon, B. J., Doucette, R., Jackson, P. C., & Diamond, J. (1984). Impulse activity evokes

precocious sprouting ofnociceptivenerves intodenervated skin. Somatosens. Res.,2, 97–126.

509. Novikov, L., Novikova, L., & Kellerth, J. O. (1997). Brain-derived neurotrophic factor

promotes axonal regeneration and long-term survival of adult rat spinal motoneurons

in vivo. Neuroscience, 79, 765–774.

510. Novikova, L., Novikov, L., & Kellerth, J. O. (1997). Effects of neurotransplants and BDNF

on the survival and regeneration of injured adult spinal motoneurons. Eur. J. Neurosci.,

9, 2774–2777.

511. Novikova, L., Novikov, L., & Kellerth, J. O. (1997). Persistent neuronal labeling by

retrograde fluorescent tracers: a comparison between Fast Blue, Fluoro-Gold and

various dextran conjugates. J. Neurosci. Methods, 74, 9–15.

512. Nukada, H. (1986). Increased susceptibility to ischemic damage in steptozocin diabetic

nerve. Diabetes, 35, 1058–1061.

513. Nukada, H. (1992). Mild ischemia causes severe pathological changes in experimental

diabetic nerve. Muscle Nerve, 15, 1116–1122.

514. Nukada, H. & Dyck, P. J. (1984). Microsphere embolization of nerve capillaries and fiber

degeneration. Am. J. Pathol., 115, 275–287.

515. Nukada, H. & Dyck, P. J. (1987). Acute ischemia causes axonal stasis, swelling,

attenuation and secondary demyelination. Ann. Neurol., 22, 311–318.

516. Nukada, H., Dyck, P. J., & Karnes, J. L. (1985). Spatial distribution of capillaries in rat

nerves: correlation to ischemic damage. Exp. Neurol., 87, 369–376.

517. Nukada, H., Powell, H. C., & Myers, R. R. (1993). Spatial distribution of nerve injury

after occlusion of individual major vessels in rat sciatic nerves. J. Neuropathol. Exp.

Neurol., 52, 452–459.

249References

518. Oblinger, M. M. & Lasek, R. J. (1984). A conditioning lesion of the peripheral axons

of dorsal root ganglion cells accelerates regeneration of only their peripheral axons.

J. Neurosci., 4, 1736–1744.

519. Oblinger, M. M., Wong, J., & Parysek, L. M. (1989). Axotomy-induced changes in the

expression of a type III neuronal intermediate filament gene. J. Neurosci., 9, 3766–3775.

520. Obrosova, I. G., Van Huysen, C., Fathallah, L. et al. (2000). Evaluation of alpha(1)-

adrenoceptor antagonist on diabetes-induced changes in peripheral nerve function,

metabolism, and antioxidative defense. FASEB J., 14, 1548–1558.

521. Ochoa, J., Fowler, T. J., & Gilliatt, R. W. (1972). Anatomical changes in peripheral nerves

compressed by a pneumatic tourniquet. J. Anat., 113, 433–455.

522. Ochoa, J. & Mair, W. G. (1969). The normal sural nerve in man. I. Ultrastructure and

numbers of fibres and cells. Acta Neuropathol. (Berl. ), 13, 197–216.

523. Ochoa, J. & Mair, W. G. (1969). The normal sural nerve in man. II. Changes in the axons

and Schwann cells due to ageing. Acta Neuropathol. (Berl.), 13, 217–239.

524. Oestreicher, A. B., De Graan, P. N. E., Gispen, W. H., Verhaagen, J., & Schrama, L. H.

(1997). B-50, the growth associated protein-43: modulation of cell morphology

and communication in the nervous system. Prog. Neurobiol. 53, 627–686.

525. Oestreicher, A. B., Van Dongen, C. J., Zwiers, H., & Gispen, W. H. (1983). Affinity-purified

anti-B-50 protein antibody: interference with the function of the phosphoprotein B-50

in synaptic plasma membranes. J. Neurochem., 41, 331–340.

526. Oestreicher, A. B., Zwiers, H., Schotman, P., & Gispen, W. H. (1981).

Immunohistochemical localization of a phosphoprotein (B-50) isolated from rat brain

synaptosomal plasma membranes. Brain Res. Bull., 6, 145–153.

527. Ohnishi, A., O’Brien, P. C., & Dyck, P. J. (1974). Studies to improve fixation of human

nerves. 2. Effect of time elapsed between death and glutaraldehyde fixation on

relationship of axonal area to number of myelin lamellae. J. Neurol. Sci., 23, 387–390.


nerves. IV. Effect of time elapsed between death and glutaraldehyde fixation on

density of microtubules and neurofilaments. J. Neuropathol. Exp. Neurol., 35, 26–29.


nerves. Part 3. Effect of osmolality of glutaraldhyde solutions on relationship of axonal

area to number of myelin lamellae. J. Neurol. Sci., 27, 193–199.


nerves. V. Effect of temperature, fixative and CaCl2 on density of microtubules and

neurofilaments. J. Neuropathol. Exp. Neurol., 35, 167–179.

531. Okajima, S., Mizoguchi, A., Tamai, K., Hirasawa, Y., & Ide, C. (1995). Distribution

of protein kinase C (alpha, beta, gamma subtypes) in normal nerve fibers and in

regenerating growth cones of the rat peripheral nervous system. Neuroscience,

66, 645–654.

532. Okajima, S., Shirasu, M., Hirasawa, Y., & Ide, C. (2000). Ultrastructural characteristics

and synaptophysin immunohistochemistry of regenerating nerve growth cones

following traumatic injury to rat peripheral nerve. J. Reconstr. Microsurg., 16, 637–642.

533. Olsson, Y. (1967). Degranulation of mast cells in peripheral nerve injuries. Acta Neurol.

Scand., 43, 365–374.

250 References

534. Olsson, Y. & Kristensson, K. (1971). Permeability of blood vessels and connective tissue

sheaths in the peripheral nervous system to exogenous proteins. Acta Neuropathol.

Suppl. 5, 61–69.

535. Olsson, Y. & Reese, T. S. (1971). Permeability of vasa nervorum and perineurium in

mouse sciatic nerve studied by fluorescence and electron microscopy.

J. Neuropathol. Exp. Neurol., 30, 105–119.

536. Ono, S. (2003). Regulation of actin filament dynamics by actin depolymerizing factor/

cofilin and actin-interacting protein 1: new blades for twisted filaments. Biochemistry,

42, 13363–13370.

537. Oppenheim, R. W., Prevette, D., Yin, Q. W., Collins, F., & MacDonald, J. (1991). Control

of embryonic motoneuron survival in vivo by ciliary neurotrophic factor. Science, 251,

1616–1618.

538. Oppenheim, R. W., Wiese, S., Prevette, D. et al. (2001). Cardiotrophin-1, a muscle-

derived cytokine, is required for the survival of subpopulations of developing

motoneurons. J. Neurosci., 21, 1283–1291.

539. Oudega, M., Xu, X. M., Guenard, V., Kleitman, N., & Bunge, M. B. (1997). A combination

of insulin-like growth factor-I and platelet-derived growth factor enhances

myelination but diminishes axonal regeneration into Schwann cell grafts in the adult

rat spinal cord. Glia, 19, 247–258.

540. Ozmen, S., Ayhan, S., Latifoglu, O., & Siemionow, M. (2002). Stamp and paper method:

a superior technique for the walking track analysis. Plast. Reconstr. Surg., 109, 1760–1761.

541. Parmantier, E., Lynn, B., Lawson, D. et al. (1999). Schwann cell-derived Desert hedgehog

controls the development of peripheral nerve sheaths. Neuron, 23, 713–724.

542. Parry, G. J. & Brown, M. J. (1981). Arachidonate-induced experimental nerve infarction.

J. Neurol. Sci., 50, 123–133.

543. Parry, G. J. & Linn, D. J. (1988). Conduction block without demyelination following

acute nerve infarction. J. Neurol. Sci., 84, 265–273.

544. Pasterkamp, R. J. & Verhaagen, J. (2006). Semaphorins in axon regeneration:

developmental guidance molecules gone wrong? Philos. Trans. R. Soc. Lond. B Biol. Sci.,

361, 1499–1511.

545. Pataky, D. M., Borisoff, J. F., Fernandes, K. J., Tetzlaff, W., & Steeves, J. D. (2000).

Fibroblast growth factor treatment produces differential effects on survival and

neurite outgrowth from identified bulbospinal neurons in vitro. Exp. Neurol., 163,

357–372.

546. Patti, M. E., Sun, X. J., Bruening, J. C. et al. (1995). 4PS/insulin receptor substrate (IRS)-2

is the alternative substrate of the insulin receptor in IRS-1-deficient mice. J. Biol. Chem.,

270, 24670–24673.

547. Paves, H. & Saarma, M. (1997). Neurotrophins as in vitro growth cone guidance

molecules for embryonic sensory neurons. Cell Tissue Res., 290, 285–297.

548. Peng, H. B., Yang, J. F., Dai, Z. et al. (2003). Differential effects of neurotrophins and

Schwann cell-derived signals on neuronal survival/growth and synaptogenesis.

J. Neurosci., 23, 5050–5060.

549. Perlson, E., Hanz, S., Ben-Yaakov, K. et al. (2005). Vimentin-dependent spatial

translocation of an activated MAP kinase in injured nerve 1. Neuron, 45, 715–726.

251References

550. Perlson, E., Medzihradszky, K. F., Darula, Z. et al. (2004). Differential proteomics reveals

multiple components in retrogradely transported axoplasm after nerve injury.

Mol. Cell Proteomics. 3, 510–520.

551. Perrin, F. E., Lacroix, S., Aviles-Trigueros, M., & David, S. (2005). Involvement of

monocyte chemoattractant protein-1, macrophage inflammatory protein-1alpha and

interleukin-1beta in Wallerian degeneration. Brain, 128, 854–866.

552. Perry, V. H., Brown, M. C., & Andersson, P. B. (1993). Macrophage responses to central

and peripheral nerve injury. Adv. Neurol., 59, 309–314.

553. Pertz, O., Hodgson, L., Klemke, R. L., & Hahn, K. M. (2006). Spatiotemporal dynamics of

RhoA activity in migrating cells. Nature, 440, 1069–1072.

554. Peters, A., Palay, S. L., & Webster, H. D. (1991). The Fine Structure of the Nervous System,

3rd edn. New York: Oxford University Press.

555. Pfister, L. A., Papaloizos, M., Merkle, H. P., & Gander, B. (2007). Nerve conduits and

growth factor delivery in peripheral nerve repair. J. Peripher. Nerv. Syst., 12, 65–82.

556. Plantinga, L. C., Verhaagen, J., Edwards, P. M. et al. (1993). The expression of B-50/GAP-43

in Schwann cells in upregulated degenerating peripheral nerve stumps following

nerve injury. Brain Res., 602, 69–76.

557. Plewnia, C., Wallace, C., & Zochodne, D. W. (1999). Traumatic sciatic neuropathy:

a novel cause, local experience, and a review of the literature. J. Trauma, 47, 986–991.

558. Plunet, W., Kwon, B. K., & Tetzlaff, W. (2002). Promoting axonal regeneration in

the central nervous system by enhancing the cell body response to axotomy.

J. Neurosci. Res., 68, 1–6.

559. Polydefkis, M., Hauer, P., Sheth, S. et al. (2004). The time course of epidermal nerve fibre

regeneration: studies in normal controls and in people with diabetes, with and

without neuropathy. Brain, 127, 1606–1615.

560. Price, R. D., Yamaji, T., Yamamoto, H. et al. (2005). FK1706, a novel non-

immunosuppressive immunophilin: neurotrophic activity and mechanism of action.

Eur. J. Pharmacol., 509, 11–19.

561. Pugliese, G., Tilton, R. G., Speedy, A. et al. (1989). Effects of very mild versus overt diabetes

on vascular haemodynamics and barrier function in rats. Diabetologia, 32, 845–857.

562. Puigdellivol-Sanchez, A., Prats-Galino, A., & Molander, C. (2006). Estimations of

topographically correct regeneration to nerve branches and skin after peripheral

nerve injury and repair. Brain Res., 1098, 49–60.

563. Qi, M. L., Wakabayashi, Y., Haro, H., & Shinomiya, K. (2003). Changes in FGF-2

expression in the distal spinal cord stump after complete cord transection:

a comparison between infant and adult rats. Spine, 28, 1934–1940.

564. Qiu, J., Cai, D., Dai, H. et al. (2002). Spinal axon regeneration induced by elevation of

cyclic AMP. Neuron, 34, 895–903.

565. Quayle, J. M., Bonev, A. D., Brayden, J. E., & Nelson, M. T. (1994). Calcitonin gene-related

peptide activated ATP-sensitive Kþ currents in rabbit arterial smooth muscle via

protein kinase Am. J. Physiol., 475 (1), 9–13.

566. Rabinovsky, E. D., Smith, G. M., Browder, D. P., Shine, H. D., & McManaman, J. L. (1992).

Peripheral nerve injury down-regulates CNTF expression in adult rat sciatic nerves.

J. Neurosci. Res., 31, 188–192.

252 References

567. Raff, M. C., Sangalang, V., & Asbury, A. K. (1968). Ischemic mononeuropathy multiplex

associated with diabetes mellitus. Arch. Neurol., 18, 487–499.

568. Rafuse, V. F., Gordon, T., & Orozco, R. (1992). Proportional enlargement of motor units

after partial denervation of cat triceps surae muscles. J. Neurophysiol., 68, 1261–1276.

569. Raivich, G., Bohatschek, M., Da Costa, C. et al. (2004). The AP-1 transcription factor c-Jun

is required for efficient axonal regeneration. Neuron, 43, 57–67.

570. Rajagopal, R., Chen, Z. Y., Lee, F. S., & Chao, M. V. (2004). Transactivation of Trk

neurotrophin receptors by G-protein-coupled receptor ligands occurs on intracellular

membranes. J. Neurosci., 24, 6650–6658.

571. Ramji, N., Toth, C., Kennedy, J., & Zochodne, D. W. (2007). Does diabetes mellitus target

motor neurons? Neurobiol. Dis., 26, 301–311.

572. Ramon y Cajal S. (1928). Degeneration and regeneration of the nervous system.

In Cajal’s Degeneration and Regeneration of the Nervous System, ed. J. DeFelipe & E. G. Jones.

Oxford: Oxford University Press.

573. Rangappa, N., Romero, A., Nelson, K. D., Eberhart, R. C., & Smith, G. M. (2000). Laminin-

coated poly(L-lactide) filaments induce robust neurite growth while providing

directional orientation. J. Biomed. Mater. Res., 51, 625–634.

574. Rechthand, E., Hervonen, A., Sato, S., & Rapoport, S. I. (1986). Distribution of

adrenergic innervation of blood vessels in peripheral nerve. Brain Res., 374, 185–189.

575. Rechthand, E. & Rapoport, S. I. (1987). Regulation of the microenvironment of

peripheral nerve: role of the blood nerve barrier. Prog. Neurobiol., 28, 303–343.

576. Rechthand, E., Sato, S., Oberg, P. A., & Rapoport, S. I. (1988). Sciatic nerve blood flow

response to carbon dioxide. Brain Res., 446, 61–66.

577. Recio-Pinto, E., Rechler, M. M., & Ishii, D. N. (1986). Effects of insulin, insulin-like

growth factor-II, and nerve growth factor on neurite formation and survival in

cultured sympathetic and sensory neurons. J. Neurosci., 6, 1211–1219.

578. Redett, R., Jari, R., Crawford, T. et al. (2005). Peripheral pathways regulate motoneuron

collateral dynamics. J. Neurosci., 25, 9406–9412.

579. Reichardt, L. F. (2006). Neurotrophin-regulated signalling pathways. Philos. Trans. R. Soc.

Lond. B Biol. Sci., 361, 1545–1564.

580. Richardson, P. M. & Lu, X. (1994). Inflammation and axonal regeneration. J. Neurol., 242,

S57–S60.

581. Richardson, P. M. & Verge, V. M. (1986). The induction of a regenerative propensity

in sensory neurons following peripheral axonal injury. J. Neurocytol., 15, 585–594.

582. Riethmacher, D., Sonnenberg-Riethmacher, E., Brinkmann, V. et al. (1997). Severe

neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature, 389,

725–730.

583. Rong, L. L., Trojaborg, W., Qu, W. et al. (2004). Antagonism of RAGE suppresses

peripheral nerve regeneration. FASEB J., 18, 1812–1817.

584. Rong, L. L., Yan, S. F., Wendt, T. et al. (2004). RAGE modulates peripheral nerve

regeneration via recruitment of both inflammatory and axonal outgrowth pathways.

FASEB J., 18, 1818–1825.

585. Rosslenbroich, V., Dai, L., Baader, S. L. et al. (2005). Collapsin response mediator protein-4

regulates F-actin bundling. Exp. Cell Res., 310, 434–444.

253References

586. Rotshenker, S. (1981). Sprouting and synapse formation by motor axons separated

from their cell bodies. Brain Res., 223, 141–145.

587. Rufer, M., Flanders, K., & Unsicker, K. (1994). Presence and regulation of transforming

growth factor beta mRNA and protein in the normal and lesioned rat sciatic nerve.

J. Neurosci. Res., 39, 412–423.

588. Rundquist, I., Smith, Q. R., Michel, M. E. et al. (1985). Sciatic nerve blood flow measured

by laser Doppler flowometry and [14C]iodoantipyrine. Am. J. Physiol., 248, H311–H317.

589. Russell, J. W., Sullivan, K. A., Windebank, A. J., Herrmann, D. N., & Feldman, E. L. (1999).

Neurons undergo apoptosis in animal and cell culture models of diabetes. Neurobiol.

Dis., 6, 347–363.

590. Sabatini, D. M., Lai, M. M., & Snyder, S. H. (1997). Neural roles of immunophilins and

their ligands. Mol. Neurobiol., 15, 223–239.

591. Sahenk, Z., Seharaseyon, J., & Mendell, J. R. (1994). CNTF potentiates peripheral nerve

regeneration. Brain Res., 655, 246–250.

592. Saito, H., Nakao, Y., Takayama, S., Toyama, Y., & Asou, H. (2005). Specific expression of

an HNK-1 carbohydrate epitope and NCAM on femoral nerve Schwann cells in mice.

Neurosci. Res., 53, 314–322.

593. Sakamoto, T., Kawazoe, Y., Shen, J. S. et al. (2003). Adenoviral gene transfer of GDNF,

BDNF and TGF beta 2, but not CNTF, cardiotrophin-1 or IGF1, protects injured adult

motoneurons after facial nerve avulsion. J. Neurosci. Res., 72, 54–64.

594. Salafsky, B., Bell, J., & Prewitt, M. A. (1968). Development of fibrillation potentials

in denervated fast and slow skeletal muscle. Am. J. Physiol., 215, 637–643.

595. Salvarezza, S. B., Lopez, H. S., & Masco, D. H. (2003). The same cellular signaling

pathways mediate survival in sensory neurons that switch their trophic requirements

during development. J. Neurochem., 85, 1347–1358.

596. Sandvig, A., Berry, M., Barrett, L. B., Butt, A., & Logan, A. (2004). Myelin-, reactive

glia-, and scar-derived CNS axon growth inhibitors: expression, receptor signaling,

and correlation with axon regeneration. Glia, 46, 225–251.

597. Sariola, H. & Saarma, M. (2003). Novel functions and signalling pathways for GDNF.

J. Cell Sci., 116, 3855–3862.

598. Sasaki, Y., Araki, T., & Milbrandt, J. (2006). Stimulation of nicotinamide adenine

dinucleotide biosynthetic pathways delays axonal degeneration after axotomy 1200.

J. Neurosci., 26, 8484–8491.

599. Saxena, S., Bucci, C., Weis, J., & Kruttgen, A. (2005). The small GTPase Rab7 controls the

endosomal trafficking and neuritogenic signaling of the nerve growth factor receptor

TrkA. J. Neurosci., 25, 10930–10940.

600. Sayan, H., Ugurlu, B., Babul, A., Take, G., & Erdogan, D. (2004). Effects of L-arginine and NG-

nitro L-arginine methyl ester on lipid peroxide, superoxide dismutase and nitrate levels

after experimental sciatic nerve ischemia–reperfusion in rats. Int J. Neurosci., 114, 349–364.

601. Scaravilli, F., Love, S., & Myers, R. (1986). X-irradiation impairs regeneration of

peripheral nerve across a gap. J. Neurocytol., 15, 439–449.

602. Scarlato, M., Ara, J., Bannerman, P., Scherer, S., & Pleasure, D. (2003). Induction of

neuropilins-1 and -2 and their ligands, Sema3A, Sema3F, and VEGF, during Wallerian

degeneration in the peripheral nervous system. Exp. Neurol., 183, 489–498.

254 References

603. Schenone, A. E. & Dyck, P. J. (1987). Which endoneurial microvessel histologic

measurements are least influenced by vasomotor tone? Brain Res., 402, 151–154.

604. Scherer, S. S. & Arroyo, E. J. (2002). Recent progress on the molecular organization of

myelinated axons. J. Peripher. Nerv. Syst., 7, 1–12.

605. Schmelzer, J. D., Zochodne, D. W., & Low, P. A. (1989). Ischemic and reperfusion injury

of rat peripheral nerve. Proc. Natl. Acad. Sci. USA, 86, 1639–1642.

606. Schmidt, R., Grabau, G., & Yip, H. (1986). Retrograde axonal transport of 125I nerve

growth factor in Ileal mesenteric nerves in vitro. Brain Res., 378, 325.

607. Schmidt, R. E. & Yip, H. K. (1985). Retrograde axonal transport in rat ileal mesenteric

nerves. Characterization using intravenously administered 125I-nerve growth factor

and effect of chemical sympathectomy. Diabetes, 34, 1222–1229.

608. Schroder, J. M. (2001). Pathology of Peripheral Nerves. Berlin: Springer.

609. Scott, S. A. (1988). Skin sensory innervation patterns in embryonic chick hindlimbs

deprived of motoneurons. Dev. Biol, 126, 362–374.

610. Seckel, B. R. (1990). Enhancement of peripheral nerve regeneration. Muscle Nerve,

13, 785–800.

611. Seidel, C. & Bicker, G. (2000). Nitric oxide and cGMP influence axonogenesis of

antennal pioneer neurons. Development, 127, 4541–4549.

612. Seijffers, R., Allchorne, A. J., & Woolf, C. J. (2006). The transcription factor ATF-3

promotes neurite outgrowth. Mol. Cell Neurosci., 32, 143–154.

613. Seilheimer, B. & Schachner, M. (1987). Regulation of neural cell adhesion molecule

expression on cultured mouse Schwann cells by nerve growth factor. EMBO J.,

6, 1611–1616.

614. Senaldi, G., Varnum, B. C., Sarmiento, U. et al. (1999). Novel neurotrophin-1/B

cell-stimulating factor-3: a cytokine of the IL-6 family. Proc. Natl. Acad. Sci. USA, 96,

11458–11463.

615. Shao, Z., Browning, J. L., Lee, X. et al. (2005). TAJ/TROY, an orphan TNF receptor family

member, binds Nogo-66 receptor 1 and regulates axonal regeneration. Neuron, 45,

353–359.

616. Sharkey, K. A. & Lomax, A. E. (2002). Structure and function of the enteric nervous

system: neurological disease and its consequences for neuromuscular function in

the gastrointestinal tract. In Neuromuscular Function and Disease, ed. W. F. Brown,

C. F. Bolton, & M. J. Aminoff. Philadelphia: W. B. Saunders, pp. 527–555.

617. Sharma, A. K. & Thomas, P. K. (1975). Peripheral nerve regeneration in experimental

diabetes. J. Neurol. Sci., 24, 417–424.

618. Sharp, P., Krishnan, M., Pullar, O. et al. (2006). Heat shock protein 27 rescues motor

neurons following nerve injury and preservesmuscle function. Exp. Neurol., 198, 511–518.

619. Shaw, G. (1998). Neurofilaments. Georgetown: Springer.

620. Shirane, M. & Nakayama, K. I. (2006). Protrudin induces neurite formation by

directional membrane trafficking. Science, 314, 818–821.

621. Shirley, D. M., Williams, S. A., & Santos, P. M. (1996). Brain-derived neurotrophic factor

and peripheral nerve regeneration: a functional evaluation. Laryngoscope, 106, 629–632.

622. Shooter, E. M. (2001). Early days of the nerve growth factor proteins. Annu. Rev. Neurosci.,

24, 601–629.

255References

623. Shy, M. E., Frohman, E. M., So, Y. T. et al. (2003). Quantitative sensory testing: report of

the Therapeutics and Technology Assessment Subcommittee of the American

Academy of Neurology. Neurology, 60, 898–904.

624. Siconolfi, L. B. & Seeds, N. W. (2001). Induction of the plasminogen activator system

accompanies peripheral nerve regeneration after sciatic nerve crush. J. Neurosci., 21,

4336–4347.

625. Siconolfi, L. B. & Seeds, N. W. (2001). Mice lacking tPA, uPA, or plasminogen genes

showed delayed functional recovery after sciatic nerve crush. J. Neurosci., 21,

4348–4355.

626. Siironen, J., Vuorinen, V., Taskinen, H. S., & Roytta, M. (1995). Axonal regeneration into

chronically denervated distal stump. 2. Active expression of type I collagen mRNA

in epineurium. Acta Neuropathol., 89, 219–226.

627. Simon, M., Porter, R., Brown, R., Coulton, G. R., & Terenghi, G. (2003). Effect of NT-4 and

BDNF delivery to damaged sciatic nerves on phenotypic recovery of fast and slow

muscles fibres. Eur. J. Neurosci., 18, 2460–2466.

628. Simon, M., Terenghi, G., Green, C. J., & Coulton, G. R. (2000). Differential effects of NT-3

on reinnervation of the fast extensor digitorum longus (EDL) and the slow soleus

muscle of rat. Eur. J. Neurosci., 12, 863–871.

629. Simpson, S. A. & Young, J. Z. (1945). Regeneration of fibre diameter after cross-unions

of visceral and somatic nerves. J. Anat., 79, 48–65.

630. Skene, J. H. & Willard, M. (1981). Axonally transported proteins associated with axon

growth in rabbit central and peripheral nervous systems. J. Cell Biol., 89, 96–103.

631. Sommer, C. & Myers, R. R. (1996). Vascular pathology in CCI neuropathy: a quantitative

temporal study. Exp. Neurol., 141, 113–119.

632. Son, Y.-J., Trachtenberg, J. T., & Thompson, W. J. (1996). Schwann cells induce and guide

sprouting and reinnervation of neuromuscular junctions. Trends Neurosci., 19, 280–285.

633. Sondell, M., Fex-Svenningsen, A., & Kanje, M. (1997). The insulin-like growth factors

I and II stimulate proliferation of different types of Schwann cells. NeuroReport, 8,

2871–2876.

634. Sondell, M., Lundborg, G., & Kanje, M. (1999). Vascular endothelial growth factor has

neurotrophic activity and stimulates axonal outgrowth, enhancing cell survival and

Schwann cell proliferation in the peripheral nervous system. J. Neurosci., 19, 5731–5740.

635. Sondell, M., Lundborg, G., & Kanje, M. (1999). Vascular endothelial growth factor

stimulates Schwann cell invasion and neovascularization of acellular nerve grafts.

Brain Res., 846, 219–228.

636. Song, H. J., Ming, G. L., & Poo, M. M. (1997). cAMP-induced switching in turning

direction of nerve growth cones. Nature, 388, 275–279.

637. Sosa, I., Reyes, O., & Kuffler, D. P. (2005). Immunosuppressants: neuroprotection and

promoting neurological recovery following peripheral nerve and spinal cord lesions.

Exp. Neurol., 195, 7–15.

638. Stalberg, E. & Trontelj, J. V. (1979). Single Fibre Electromyography, 1st edn. Old Woking,

Surrey: Miravelle Press.

639. Steinmetz, M. P., Horn, K. P., Tom, V. J. et al. (2005). Chronic enhancement of the

intrinsic growth capacity of sensory neurons combined with the degradation of

256 References

inhibitory proteoglycans allows functional regeneration of sensory axons through the

dorsal root entry zone in the mammalian spinal cord. J. Neurosci., 25, 8066–8076.

640. Sterne, G. D., Brown, R. A., Green, C. J., & Terenghi, G. (1997). Neurotrophin-3 delivered

locally via fibronectin mats enhances peripheral nerve regeneration. Eur. J. Neurosci., 9,

1388–1396.

641. Sterne, G. D., Coulton, G. R., Brown, R. A., Green, C. J., & Terenghi, G. (1997).

Neurotrophin-3-enhanced nerve regeneration selectively improves recovery of muscle

fibers expressing myosin heavy chains 2b. J. Cell Biol., 139, 709–715.

642. Stevens, M. J., Obrosova, I., Cao, X., Van Huysen, C., & Greene, D. A. (2000). Effects of

DL-alpha-lipoic acid on peripheral nerve conduction, blood flow, energy metabolism,

and oxidative stress in experimental diabetic neuropathy. Diabetes, 49, 1006–1015.

643. Stewart, H. J., Eccleston, P. A., Jessen, K. R., & Mirsky, R. (1991). Interaction between

cAMP elevation, identified growth factors, and serum components in regulating

Schwann cell growth. J. Neurosci. Res., 30, 346–352.

644. Stewart, H. J., Turner, D., Jessen, K. R., & Mirsky, R. (1997). Expression and regulation of

alpha1beta1 integrin in Schwann cells. J. Neurobiol., 33, 914–928.

645. Stoll, G., Griffin, J. W., Li, C. Y., & Trapp, B. D. (1989). Wallerian degeneration in the

peripheral nervous system: participation of both Schwann cells and macrophages in

myelin degradation. J. Neurocytol., 18, 671–683.

646. Stoll, G. & Jander, S. (1999). The role of microglia and macrophages in the

pathophysiology of the CNS. Prog. Neurobiol., 58, 233–247.

647. Stoll, G., Jander, S., & Myers, R. R. (2002). Degeneration and regeneration of the

peripheral nervous system: from Augustus Waller’s observations to

neuroinflammation. J. Peripher. Nerv. Syst., 7, 13–27.

648. Stoll, G., Jung, S., Jander, S., van der, M. P., & Hartung, H. P. (1993). Tumor necrosis

factor-alpha in immune-mediated demyelination and Wallerian degeneration of the

rat peripheral nervous system. J. Neuroimmunol., 45, 175–182.

649. Storkebaum, E. & Carmeliet, P. (2004). VEGF: a critical player in neurodegeneration.

J. Clin. Invest., 113, 14–18.

650. Strand, F. L. & Kung, T. T. (1980). ACTH accelerates recovery of neuromuscular function

following crushing of peripheral nerve. Peptides, 1, 135–138.

651. Streppel, M., Azzolin, N., Dohm, S. et al. (2002). Focal application of neutralizing

antibodies to soluble neurotrophic factors reduces collateral axonal branching after

peripheral nerve lesion. Eur. J. Neurosci., 15, 1327–1342.

652. Sugimoto, H., Monafo, W. W., & Eliasson, S. G. (1986). Regional sciatic nerve and muscle

blood flow in conscious and anesthetized rats. Am. J. Physiol., 251, H1211–H1216.

653. Sugimoto, K., Murakawa, Y., Zhang, W., Xu, G., & Sima, A. A. (2000). Insulin receptor in

rat peripheral nerve: its localization and alternatively spliced isoforms. Diabetes Metab.

Res. Rev., 16, 354–363.

654. Sugimoto, K. & Yagihashi, S. (1997). Effects of aminoguanidine on structural

alterations of microvessels in peripheral nerve of streptozotocin diabetic rats.

Microvasc. Res., 53, 105–112.

655. Sugiura, S., Lahav, R., Han, J. et al. (2000). Leukaemia inhibitory factor is required

for normal inflammatory responses to injury in the peripheral and central

257References

nervous systems in vivo and is chemotactic for macrophages in vitro. Eur. J. Neurosci.,

12, 457–466.

656. Sulaiman, O. A. & Gordon, T. (2000). Effects of short- and long-term Schwann cell

denervation on peripheral nerve regeneration, myelination, and size. Glia, 32, 234–246.

657. Sulaiman, O. A. & Gordon, T. (2002). Transforming growth factor-beta and forskolin

attenuate the adverse effects of long-term Schwann cell denervation on peripheral

nerve regeneration in vivo. Glia, 37, 206–218.

658. Sulaiman, O. A., Midha, R., Munro, C. A. et al. (2002). Chronic Schwann cell denervation

and the presence of a sensory nerve reduce motor axonal regeneration. Exp. Neurol.,

176, 342–354.

659. Sulaiman, O. A., Voda, J., Gold, B. G., & Gordon, T. (2002). FK506 increases peripheral

nerve regeneration after chronic axotomy but not after chronic Schwann cell

denervation. Exp. Neurol., 175, 127–137.

660. Sunderland, S. (1978). Nerves and Nerve Injuries, 2nd edn. Edinburgh: Churchill

Livingstone.

661. Suter, U. & Martini, R. (2005). Myelination. In Peripheral Neuropathy, 4th edn., ed.

P. J. Dyck & P. K. Thomas. Philadelphia: Elsevier Saunders, pp. 411–431.

662. Sutera, S. P., Chang, K., Marvel, J., & Williamson, J. R. (1992). Concurrent increases

in regional hematocrit and blood flow in diabetic rats: prevention by sorbinil.

Am. J. Physiol., 263, H945–H950.

663. Tam, S. L., Archibald, V., Jassar, B., Tyreman, N., & Gordon, T. (2001). Increased

neuromuscular activity reduces sprouting in partially denervated muscles. J. Neurosci.,

21, 654–667.

664. Tam, S. L. & Gordon, T. (2003). Mechanisms controlling axonal sprouting at the

neuromuscular junction. J. Neurocytol., 32, 961–974.

665. Tan, W., Rouen, S., Barkus, K. M. et al. (2003). Nerve growth factor blocks the glucose-

induced down-regulation of caveolin-1 expression in Schwann cells via p75

neurotrophin receptor signaling. J. Biol. Chem., 278, 23151–23162.

666. Tanaka, A., Kamiakito, T., Hakamata, Y. et al. (2001). Extensive neuronal localization

and neurotrophic function of fibroblast growth factor 8 in the nervous system.

Brain Res., 912, 105–115.

667. Tanaka, T., Saito, H., & Matsuki, N. (1993). Endogenous nitric oxide inhibits NMDA-

and kainate-responses by a negative feedback system in rat hippocampal neurons.

Brain Res., 631, 72–76.

668. Tang, F. & Kalil, K. (2005). Netrin-1 induces axon branching in developing cortical

neurons by frequency-dependent calcium signaling pathways. J. Neurosci., 25, 6702–6715.

669. Tani, T., Miyamoto, Y., Fujimori, K. E. et al. (2005). Trafficking of a ligand-receptor

complex on the growth cones as an essential step for the uptake of nerve growth factor

at the distal end of the axon: a single-molecule analysis. J. Neurosci., 25, 2181–2191.

670. Teng, F. Y. & Tang, B. L. (2006). Axonal regeneration in adult CNS neurons – signaling

molecules and pathways. J. Neurochem., 96, 1501–1508.

671. Terada, M., Yasuda, H., & Kikkawa, R. (1998). Delayed Wallerian degeneration and

increased neurofilament phosphorylation in sciatic nerves of rats with streptozocin-

induced diabetes. J. Neurol. Sci., 155, 23–30.

258 References

672. Terada, M., Yasuda, H., Kikkawa, R., & Shigeta, Y. (1996). Tolrestat improves nerve

regeneration after crush injury in streptozocin-induced diabetic rats. Metabolism,

45, 1189–1195.

673. Terashima, T., Yasuda, H., Terada, M. et al. (2001). Expression of Rho-family GTPases

(Rac, cdc42, RhoA) and their association with p-21 activated kinase in adult rat

peripheral nerve. J. Neurochem., 77, 986–993.

674. Tesfaye, S., Harris, N., Jakubowski, J. J. et al. (1993). Impaired blood flow and arterior-

venous shunting in human diabetic neuropathy: a novel technique of nerve

photography and fluorescein angiography. Diabetologia, 36, 1266–1274.

675. Tetzlaff, W., Bisby, M. A., & Kreutzberg, G. W. (1988). Changes in cytoskeletal proteins

in the rat facial nucleus following axotomy. J. Neurosci., 8, 3181–3189.

676. Tetzlaff, W., Zwiers, H., Lederis, K., Cassar, L., & Bisby, M. A. (1989). Axonal transport

and localization of B-50/GAP-43-like immunoreactivity in regenerating sciatic and

facial nerves of the rat. J. Neurosci., 9, 1303–1313.

677. Theriault, E. & Diamond, J. (1988). Nociceptive cutaneous stimuli evoke localized

contractions in a skeletal muscle. J. Neurophysiol., 60, 446–462.

678. Theriault,M., Dort, J., Sutherland,G., & Zochodne, D.W. (1998). A prospective quantitative

study of sensory deficits after whole sural nerve biopsies in diabetic and nondiabetic

patients. Surgical approach and the role of collateral sprouting. Neurology, 50, 480–484.

679. Thoenen, H. & Barde, Y. A. (1980). Physiology of nerve growth factor. Physiol. Rev.,

60, 1284–1335.

680. Tilton, R. G., Chang, K., Nyengaard, J. R. et al. (1995). Inhibition of sorbitol

dehydrogenase. Effects on vascular and neural dysfunction in streptozocin-induced

diabetic rats. Diabetes, 44, 234–242.

681. Tinel, J. (1918). Nerve Wounds, Symptomatology of Peripheral Nerve Lesions Caused by War

Wounds. New York: William Wood and Company.

682. Tofaris, G. K., Patterson, P. H., Jessen, K. R., & Mirsky, R. (2002). Denervated Schwann

cells attract macrophages by secretion of leukemia inhibitory factor (LIF) and

monocyte chemoattractant protein-1 in a process regulated by interleukin-6 and LIF.

J. Neurosci., 22, 6696–6703.

683. Tolwani, R. J., Cosgaya, J. M., Varma, S. et al. (2004). BDNF overexpression produces

a long-term increase in myelin formation in the peripheral nervous system. J. Neurosci.

Res., 77, 662–669.

684. Tonge, D., Edbladh, M., Edstrom, A., & Kanje, M. (1992). An improved HRP method

for tracing axons in whole-mount preparations during early stages of regeneration

in peripheral nerves of adult animals. J. Neurosci. Methods, 44, 27–31.

685. Tonge, D., Edstrom, A., & Ekstrom, P. (1998). Use of explant cultures of peripheral

nerves of adult vertebrates to study axonal regeneration in vitro. Prog. Neurobiol.,

54, 459–480.

686. Tonra, J. R., Curtis, R., Wong, V. et al. (1998). Axotomy upregulates the anterograde

transport and expression of brain-derived neurotrophic factor by sensory neurons.

J. Neurosci., 18, 4374–4383.

687. Torigoe, K., Hashimoto, K., & Lundborg, G. (1999). A role of migratory Schwann cells in

a conditioning effect of peripheral nerve regeneration. Exp. Neurol., 160, 99–108.

259References

688. Tosney, K. W. & Landmesser, L. T. (1985). Growth cone morphology and trajectory in

the lumbosacral region of the chick embryo. J. Neurosci., 5, 2345–2358.

689. Toth, C., Brussee, V., & Martinez, J. A. et al. (2006). Rescue and regeneration of injured

peripheral nerve axons by intrathecal insulin. Neuroscience, 139, 429–449.

690. Toth, C., Brussee, V., & Zochodne, D. W. (2006). Remote neurotrophic support of

epidermal nerve fibres in experimental diabetes. Diabetologia, 49, 1081–1088.

691. Toth, C., Martinez, J. A., & Zochodne, D. W. (2006). Local axon synthesis of CGRP peptide

is essential for adult nerve regeneration [abstract]. Soc. Neurosci. Abs., 2006.

692. Triban, C., Guidolin, D., Fabris, M. et al. (1989). Ganglioside treatment and improved

axonal regeneration capacity in experimental diabetic neuropathy. Diabetes, 38,

1012–1022.

693. Triolo, D., Dina, G., Lorenzetti, I. et al. (2006). Loss of glial fibrillary acidic protein

(GFAP) impairs Schwann cell proliferation and delays nerve regeneration after

damage. J. Cell Sci., 119, 3981–3993.

694. Truong, W., Cheng, C., Xu, Q. G., Li, X. Q., & Zochodne, D. W. (2003). Mu opioid

receptors and analgesia at the site of a peripheral nerve injury. Ann. Neurol., 53,

366–375.

695. Tseng, T. J., Chen, C. C., Hsieh, Y. L., & Hsieh, S. T. (2007). Effects of decompression

on neuropathic pain behaviors and skin reinnervation in chronic constriction injury.

Exp. Neurol., 204, 574–582.

696. Tsuda, M., Shigemoto-Mogami, Y., Koizumi, S. et al. (2003). P2X4 receptors induced in

spinal microglia gate tactile allodynia after nerve injury. Nature, 424, 778–783.

697. Tuck, R. R., Schmelzer, J. D., & Low, P. A. (1984). Endoneurial blood flow and oxygen

tension in the sciatic nerves of rats with experimental diabetic neuropathy. Brain, 107,

935–950.

698. Tucker, B. A., Rahimtula, M., & Mearow, K. M. (2006). Laminin and growth factor

receptor activation stimulates differential growth responses in subpopulations of

adult DRG neurons. Eur. J. Neurosci., 24, 676–690.

699. Tucker, K. L., Meyer, M., & Barde, Y. A. (2001). Neurotrophins are required for nerve

growth during development. Nat. Neurosci., 4, 29–37.

700. Udina, E., Voda, J., Gold, B. G., & Navarro, X. (2003). Comparative dose-dependence

study of FK506 on transected mouse sciatic nerve repaired by allograft or xenograft.

J. Peripher. Nerv. Syst., 8, 145–154.

701. Valdez, G., Akmentin, W., Philippidou, P. et al. (2005). Pincher-mediated

macroendocytosis underlies retrograde signaling by neurotrophin receptors.

J. Neurosci., 25, 5236–5247.

702. van Kesteren, R. E., Carter, C., Dissel, H. M. et al. (2006). Local synthesis of actin-binding

protein beta-thymosin regulates neurite outgrowth. J. Neurosci., 26, 152–157.

703. van Minnen, J., Bergman, J. J., Van Kesteren, E. R. et al. (1997). De novo protein synthesis

in isolated axons of identified neurons. Neuroscience, 80, 1–7.

704. Van, D. V. & Roberts, L. J. (1999). Contrasting roles for nitric oxide and peroxynitrite in

the peroxidation of myelin lipids. J. Neuroimmunol., 95, 1–7.

705. Vance, J. E., Campenot, R. B., & Vance, D. E. (2000). The synthesis and transport of lipids

for axonal growth and nerve regeneration. Biochim. Biophys. Acta, 1486, 84–96.

260 References

706. Verdu, E., Ceballos, D., Vilches, J. J., & Navarro, X. (2000). Influence of aging on

peripheral nerve function and regeneration. J. Peripher. Nerv. Syst., 5, 191–208.

707. Verge, V. M. K., Gratto, K. A., Karchewski, L. A., & Richardson, P. M. (1996).

Neurotrophins and nerve injury in the adult. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 351,

423–430.

708. Verge, V. M. K., Tetzlaff, W., Bisby, M. A., & Richardson, P. M. (1990). Influence of nerve

growth factor on neurofilament gene expression in mature primary sensory neurons.

J. Neurosci., 10, 2018–2025.

709. Verhaagen, J., van Hooff, C. O., Edwards, P. M. et al. (1986). The kinase C substrate

protein B-50 and axonal regeneration. Brain Res. Bull., 17, 737–741.

710. Verma, P., Chierzi, S., Codd, A. M. et al. (2005). Axonal protein synthesis and degradation

are necessary for efficient growth cone regeneration. J. Neurosci., 25, 331–342.

711. Visochek, L., Steingart, R. A., Vulih-Shultzman, I. et al. (2005). PolyADP-ribosylation is

involved in neurotrophic activity. J. Neurosci., 25, 7420–7428.

712. Vita, G., Dattola, R., Girlanda, P. et al. (1983). Effects of steroid hormones on muscle

reinnervation after nerve crush in rabbit. Exp. Neurol., 80, 279–287.

713. Wagner, R. & Myers, R. R. (1996). Schwann cells produce tumor necrosis factor alpha:

expression in injured and non-injured nerves. Neuroscience, 73, 625–629.

714. Wall, J. T., Kaas, J. H., Sur, M. et al. (1986). Functional reorganization in somatosensory

cortical areas 3b and 1 of adult monkeys after median nerve repair: possible

relationships to sensory recovery in humans. J. Neurosci., 6, 218–233.

715. Waller, A. (1850). Experiments on the section of the glossopharyngeal and hypoglossal

nerves of the frog and observations of the alterations produced thereby in the

structure of their primitive fibers. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 140, 423–429.

716. Wallquist, W., Plantman, S., Thams, S. et al. (2005). Impeded interaction between

Schwann cells and axons in the absence of laminin alpha4. J. Neurosci., 25, 3692–3700.

717. Wallquist, W., Zelano, J., Plantman, S. et al. (2004). Dorsal root ganglion neurons

up-regulate the expression of laminin-associated integrins after peripheral but not

central axotomy. J. Comp. Neurol., 480, 162–169.

718. Walsh, F. S. & Doherty, P. (1996). Cell adhesion molecules and neuronal regeneration.

Curr. Opin. Cell Biol., 8, 707–713.

719. Walsh, G. S., Orike, N., Kaplan, D. R., & Miller, F. D. (2004). The invulnerability of adult

neurons: a critical role for p73. J. Neurosci., 24, 9638–9647.

720. Wang, G. & Scott, S. A. (1999). Independent development of sensory and motor

innervation patterns in embryonic chick hindlimbs. Dev. Biol., 208, 324–336.

721. Wang, M. S., Zeleny-Pooley, M., & Gold, B. G. (1997). Comparative dose-dependence

study of FK506 and cyclosporin A on the rate of axonal regeneration in the rat sciatic

nerve. J. Pharmacol. Exp. Ther., 282, 1084–1093.

722. Wang, X., Wang, C., Zeng, J. et al. (2005). Gene transfer to dorsal root ganglia by

intrathecal injection: effects on regeneration of peripheral nerves. Mol. Ther., 12,

314–320.

723. Wang, Y., Shibasaki, F., & Mizuno, K. (2005). Calcium signal-induced cofilin

dephosphorylation is mediated by Slingshot via calcineurin. J. Biol. Chem., 280,

12683–12689.

261References

724. Wang, Y. L., Wang, D. Z., & Nie, X. et al. (2006). The role of bone morphogenetic protein-2

in vivo in regeneration of peripheral nerves. Br. J. Oral Maxillofac. Surg., 45, 197–202.

725. Wanner, I. B., Mahoney, J., Jessen, K. R. et al. (2006). Invariantmantling of growth cones by

Schwann cell precursors characterize growing peripheral nerve fronts. Glia, 54, 424–438.

726. Watson, F. L., Heerssen, H. M., Bhattacharyya, A. et al. (2001). Neurotrophins use the

Erk5 pathway to mediate a retrograde survival response. Nat. Neurosci., 4, 981–988.

727. Watson, F. L., Heerssen, H. M., Moheban, D. B. et al. (1999). Rapid nuclear responses to

target-derived neurotrophins require retrograde transport of ligand-receptor complex.

J. Neurosci., 19, 7889–7900.

728. Watts, R. J., Hoopfer, E. D., & Luo, L. (2003). Axon pruning during Drosophila

metamorphosis: evidence for local degeneration and requirement of the

ubiquitin–proteasome system. Neuron, 38, 871–885.

729. Wayman, G. A., Kaech, S., Grant, W. F. et al. (2004). Regulation of axonal extension

and growth cone motility by calmodulin-dependent protein kinase I. J. Neurosci., 24,

3786–3794.

730. Weerasuriya, A. (1987). Permeability of endoneurial capillaries to K, Na, and Cl and its

relation to peripheral nerve excitability. Brain Res., 419, 188–196.

731. Weerasuriya, A. (1988). Patterns of change in endoneurial capillary permeability and

vascular space during Wallerian degeneration. Brain Res., 445, 181–187.

732. Weerasuriya, A. (1990). Patterns of change in endoneurial capillary permeability and

vascular space during nerve regeneration. Brain Res., 510, 135–139.

733. Weerasuriya, A., Curran, G. L., & Poduslo, J. F. (1989). Blood-nerve transfer of albumin

and its implications for the endoneurial microenvironment. Brain Res., 494, 114–121.

734. Weerasuriya, A., Rapoport, S. I., & Taylor, R. E. (1980). Ionic permeabilities of the frog

perineurium. Brain Res., 191, 405–415.

735. Wehrman, T., He, X., Raab, B. et al. (2007). Structural and mechanistic insights into

nerve growth factor interactions with the TrkA and p75 receptors. Neuron, 53, 25–38.

736. Werner, A., Willem, M., Jones, L. L. et al. (2000). Impaired axonal regeneration in alpha7

integrin-deficient mice. J. Neurosci., 20, 1822–1830.

737. White, M. F. (1997). The insulin signalling system and the IRS proteins. Diabetologia,

40, S2–S17.

738. White, M. F. & Yenush, L. (1998). The IRS-signaling system: a network of docking proteins

that mediate insulin and cytokine action. Curr. Top. Microbiol. Immunol., 228, 179–208.

739. Wiberg, M., Hazari, A., Ljungberg, C. et al. (2003). Sensory recovery after hand

reimplantation: a clinical, morphological, and neurophysiological study in humans.

Scand. J. Plast. Reconstr. Surg. Hand Surg., 37, 163–173.

740. Williams, L. R., Longo, F. M., Powell, H. C., Lundborg, G., & Varon, S. (1983). Spatial-

temporal progress of peripheral nerve regeneration within a silicone chamber:

parameters for a bioassay. J. Comp. Neurol., 218, 460–470.

741. Willis, D. & Coggeshall, R. (1991). Sensory Mechanisms of the Spinal Cord, 2nd edn.

New York: Plenum.

742. Willis, D., Li, K. W., Zheng, J. Q. et al. (2005). Differential transport and local translation

of cytoskeletal, injury-response, and neurodegeneration protein mRNAs in axons.

J. Neurosci., 25, 778–791.

262 References

743. Willis W. D. & Coggeshall R. E. (2004). Sensory Mechanisms of the Spinal Cord, 3rd edn. New

York: Kluwer Academic/Plenum.

744. Witzel, C., Rohde, C., & Brushart, T. M. (2005). Pathway sampling by regenerating

peripheral axons. J. Comp. Neurol., 485, 183–190.

745. Woldeyesus, M. T., Britsch, S., Riethmacher, D. et al. (1999). Peripheral nervous system

defects in erbB2 mutants following genetic rescue of heart development. Genes Dev.,

13, 2538–2548.

746. Wong, L. F., Yip, P. K., Battaglia, A. et al. (2006). Retinoic acid receptor beta2

promotes functional regeneration of sensory axons in the spinal cord. Nat. Neurosci.,

9, 243–250.

747. Woo, S. & Gomez, T. M. (2006). Rac1 and RhoA promote neurite outgrowth through

formation and stabilization of growth cone point contacts. J. Neurosci., 26, 1418–1428.

748. Wood, J. N., Boorman, J. P., Okuse, K., & Baker, M. D. (2004). Voltage-gated sodium

channels and pain pathways. J. Neurobiol., 61, 55–71.

749. Wu, K. Y., Hengst, U., Cox, L. J. et al. (2005). Local translation of RhoA regulates growth

cone collapse. Nature, 436, 1020–1024.

750. Wu, L. C., Goettl, V. M., Madiai, F., Hackshaw, K. V., & Hussain, S. R. (2006). Reciprocal

regulation of nuclear factor kappa B and its inhibitor ZAS3 after peripheral nerve

injury. BMC. Neurosci. 7, 4.

751. Xu, G., Pierson, C. R., Murakawa, Y., & Sima, A. A. (2002). Altered tubulin and

neurofilament expression and impaired axonal growth in diabetic nerve regeneration.

J. Neuropathol. Exp. Neurol., 61, 164–175.

752. Xu, G. & Sima, A. A. (2001). Altered immediate early gene expression in injured

diabetic nerve: implications in regeneration. J. Neuropathol. Exp. Neurol., 60, 972–983.

753. Xu, Q.-G., Cheng, C., Sun, H., Thomsen, K., & Zochodne, D. W. (2003). Local sensory

ganglion ischemia induced by endothelin vasoconstriction. Neuroscience, 122, 897–905.

754. Xu, Q. G., Li, X.-Q., Kotecha, S. A. et al. (2004). Insulin as an in vivo growth factor.

Exp. Neurol., 188, 43–51.

755. Xu, Q. G. & Zochodne, D. W. (2002). Ischemia and failed regeneration in chronic

experimental neuromas. Brain Res., 946, 24–30.

756. Yamamoto, Y., Yasuda, Y., Kimura, Y., & Komiya, Y. (1998). Effects of cilostazol, an

antiplatelet agent, on axonal regeneration following nerve injury in diabetic rats.

Eur. J. Pharmacol., 352, 171–178.

757. Yamasaki, H., Itakura, C., & Mizutani, M. (1991). Hereditary hypotrophic axonopathy

with neurofilament deficiency in a mutant strain of the Japanese quail. Acta

Neuropathol. (Berl.), 82, 427–434.

758. Yamashita, T. & Tohyama, M. (2003). The p75 receptor acts as a displacement factor

that releases Rho from Rho-GDI. Nat. Neurosci., 6, 461–467.

759. Yao, M., Inserra, M. M., Duh, M. J., & Terris, D. J. (1998). A longitudinal, functional study

of peripheral nerve recovery in the mouse. Laryngoscope, 108, 1141–1145.

760. Yao, M., Moir, M. S., Wang, M. Z., To, M. P., & Terris, D. J. (1999). Peripheral nerve

regeneration in CNTF knockout mice. Laryngoscope, 109, 1263–1268.

761. Yarmola, E. G. & Bubb, M. R. (2006). Profilin: emerging concepts and lingering

misconceptions. Trends Biochem. Sci., 31, 197–205.

263References

762. Yasuda, H., Terada, M., Maeda, K. et al. (2003). Diabetic neuropathy and nerve

regeneration. Prog. Neurobiol., 69, 229–285.

763. Yasuda, H., Terada, M., Taniguchi, Y. et al. (1999). Impaired regeneration and no

amelioration with aldose reductase inhibitor in crushed unmyelinated nerve fibers

of diabetic rats. NeuroReport, 10, 2405–2409.

764. Ygge, J. (1989). Neuronal loss in lumbar dorsal root ganglia after proximal compared to

distal sciatic nerve resection: a quantitative study in the rat. Brain Res., 478, 193–195.

765. Yin, Q., Kemp, G. J., Yu, L. G., Wagstaff, S. C., & Frostick, S. P. (2001). Neurotrophin-4

delivered by fibrin glue promotes peripheral nerve regeneration. Muscle Nerve, 24,

345–351.

766. Yin, Y., Henzl, M. T., Lorber, B. et al. (2006). Oncomodulin is a macrophage-derived

signal for axon regeneration in retinal ganglion cells. Nat. Neurosci., 9, 843–852.

767. Yosipovitch, G., Yarnitsky, D., Mermelstein, V. et al. (1995). Paradoxical heat sensation

in uremic polyneuropathy. Muscle Nerve, 18, 768–771.

768. You, S., Petrov, T., Chung, P. H., & Gordon, T. (1997). The expression of the low affinity

nerve growth factor receptor in long-term denervated Schwann cells. Glia, 20, 87–100.

769. Yuan, X. B., Jin, M., Xu, X. et al. (2003). Signalling and crosstalk of Rho GTPases in

mediating axon guidance. Nat. Cell Biol., 5, 38–45.

770. Zelena, J. (1984). Multiple axon terminals in reinnervated Pacinian corpuscles of adult

rat. J. Neurocytol., 13, 665–684.

771. Zhai, Q., Wang, J., Kim, A. et al. (2003). Involvement of the ubiquitin-proteasome system

in the early stages of Wallerian degeneration. Neuron, 39, 217–225.

772. Zhang, J. Y., Luo, X. G., Xian, C. J., Liu, Z. H., & Zhou, X. F. (2000). Endogenous BDNF

is required for myelination and regeneration of injured sciatic nerve in rodents.

Eur. J. Neurosci., 12, 4171–4180.

773. Zhang, Q. L., Liu, J., Lin, P. X., & Webster, H. (2002). Local administration of vasoactive

intestinal peptide after nerve transection accelerates early myelination and growth

of regenerating axons. J. Peripher. Nerv. Syst., 7, 118–127.

774. Zhang, X. F., Schaefer, A. W., Burnette, D. T., Schoonderwoert, V. T., & Forscher, P.

(2003). Rho-dependent contractile responses in the neuronal growth cone are

independent of classical peripheral retrograde actin flow. Neuron, 40, 931–944.

775. Zhao, Q., Dahlin, L. B., Kanje, M., & Lundborg, G. (1992). Specificity of muscle

reinnervation following repair of the transected sciatic nerve. A comparative study

of different repair techniques in the rat. J. Hand Surg. [Br]., 17, 257–261.

776. Zhao, Q., Dahlin, L. B., Kanje, M., & Lundborg, G. (1992). The formation of a

“pseudo-nerve” in silicone chambers in the absence of regenerating axons. Brain Res.,

592, 106–114.

777. Zhao, Z., Alam, S., Oppenheim, R. W. et al. (2004). Overexpression of glial cell

line-derived neurotrophic factor in the CNS rescues motoneurons from programmed

cell death and promotes their long-term survival following axotomy. Exp. Neurol.,

190, 356–372.

778. Zhong, J., Dietzel, I. D., Wahle, P., Kopf, M., & Heumann, R. (1999). Sensory impairments

and delayed regeneration of sensory axons in interleukin-6-deficient mice. J. Neurosci.,

19, 4305–4313.

264 References

779. Zhou, F. Q. & Snider, W. D. (2006). Intracellular control of developmental and

regenerative axon growth. Phil. Trans. R. Soc. Lond. B Biol. Sci., 361, 1575–1592.

780. Zhou, F. Q., Walzer, M., Wu, Y. H. et al. (2006). Neurotrophins support regenerative axon

assembly over CSPGs by an ECM-integrin-independent mechanism. J. Cell Sci., 119,

2787–2796.

781. Zigmond, R. E. (1997). LIF, NGF, and the cell body response to axotomy. Neuroscientist,

3, 176–185.

782. Zochodne, D. W. (1999). Nerve growth factor. Science and Medicine, 6, 46–55.

783. Zochodne, D. W. (2002). Nerve and ganglion blood flow in diabetes: an appraisal.

In Neurobiology of Diabetic Neuropathy, ed. D. Tomlinson. San Diego: Academic Press,

pp. 161–202.

784. Zochodne, D. W., Allison, J. A., Ho, W. et al. (1995). Evidence for CGRP accumulation and

activity in experimental neuromas. Am. J. Physiol., 268, H584–H590.

785. Zochodne, D. W., Bolton, C. F., Wells, G. A. et al. (1987). Critical illness polyneuropathy

a complication of sepsis and multiple organ failure. Brain, 110, 819–842.

786. Zochodne, D. W. & Cheng, C. (1999). Diabetic peripheral nerves are susceptible to

multifocal ischemic damage from endothelin. Brain Res., 838, 11–17.

787. Zochodne, D. W. & Cheng, C. (2000). Neurotrophins and other growth factors in the

regenerative milieu of proximal nerve stump tips. J. Anat., 196, 279–283.

788. Zochodne, D. W., Cheng, C., Miampamba, M., Hargreaves, K., & Sharkey, K. A. (2001).

Peptide accumulations in proximal endbulbs of transected axons. Brain Res., 902,

40–50.

789. Zochodne, D. W., Cheng, C., & Sun, H. (1996). Diabetes increases sciatic nerve

susceptibility to endothelin-induced ischemia. Diabetes, 45, 627–632.

790. Zochodne, D. W. & Ho, L. T. (1990). Endoneurial microenvironment and acute nerve

crush injury in the rat sciatic nerve. Brain Res., 535, 43–48.

791. Zochodne, D. W. & Ho, L. T. (1991). Influence of perivascular peptides on endoneurial

blood flow and microvascular resistance in the sciatic nerve of the rat. J. Physiol., 444,

615–630.

792. Zochodne, D. W. & Ho, L. T. (1991). Unique microvascular characteristics of the dorsal

root ganglion in the rat. Brain Res., 559, 89–93.

793. Zochodne, D. W. & Ho, L. T. (1992). Hyperemia of injured peripheral nerve: sensitivity

to CGRP antagonism. Brain Res., 598, 59–66.

794. Zochodne, D. W. & Ho, L. T. (1992). Normal blood flow but lower oxygen tension in

diabetes of young rats: microenvironment and the influence of sympathectomy.

Can. J. Physiol. Pharmacol., 70, 651–659.

795. Zochodne, D. W. & Ho, L. T. (1993). Vasa nervorum constriction from substance P

and calcitonin gene-related peptide antagonists: sensitivity to phentolamine and

nimodipine. Regul. Pept., 47, 285–290.

796. Zochodne, D. W. & Ho, L. T. (1994). Neonatal guanethidine treatment alters

endoneurial but not dorsal root ganglion perfusion in the rat. Brain Res., 649, 147–150.

797. Zochodne, D. W., Ho, L. T., & Gross, P. M. (1992). Acute endoneurial ischemia

induced by epineurial endothelin in the rat sciatic nerve. Am. J. Physiol., 263,

H1806–H1810.

265References

798. Zochodne, D. W., Huang, Z. X., Ward, K. K., & Low, P. A. (1990). Guanethidine-induced

adrenergic sympathectomy augments endoneurial perfusion and lowers endoneurial

microvascular resistance. Brain Res., 519, 112–117.

799. Zochodne, D. W. & Levy, D. (2005). Nitric oxide in damage, disease and repair of the

peripheral nervous system. Cell Mol. Biol. (Noisy.-le-grand), 51, 255–267.

800. Zochodne, D. W., Levy, D., Zwiers, H. et al. (1999). Evidence for nitric oxide and nitric

oxide synthase activity in proximal stumps of transected peripheral nerves.


801. Zochodne, D. W. & Low, P. A. (1990). Adrenergic control of nerve blood flow.

Exp. Neurol., 109, 300–307.

802. Zochodne, D. W., Low, P. A., & Dyck, P. J. (1989). Adrenergic sympathectomy ablates

unmyelinated fibers in the rat “preganglionic” cervical sympathetic trunk. Brain Res.,

498, 221–228.

803. Zochodne, D. W., Misra, M., Cheng, C., & Sun, H. (1997). Inhibition of nitric oxide

synthase enhances peripheral nerve regeneration in mice. Neurosci. Lett., 228, 71–74.

804. Zochodne, D. W., Murray, M. M., van der Sloot, P., & Riopelle, R. J. (1995). Distal tibial

mononeuropathy in diabetic and nondiabetic rats reared on wire cages: an

experimental entrapment neuropathy. Brain Res., 698, 130–136.

805. Zochodne, D. W. & Nguyen, C. (1997). Angiogenesis at the site of neuroma formation in

transected peripheral nerve. J. Anat., 191, 23–30.

806. Zochodne, D. W. & Nguyen, C. (1999). Increased peripheral nerve microvessels in early

experimental diabetic neuropathy: quantitative studies of nerve and dorsal root

ganglia. J. Neurol. Sci., 166, 40–46.

807. Zochodne, D. W., Nguyen, C., & Sharkey, K. A. (1994). Accumulation and degranulation

of mast cells in experimental neuromas. Neurosci. Lett., 182, 3–6.

808. Zochodne, D. W., Sun, H., & Li, X. Q. (2001). Evidence that nitric oxide and opioid

containing inerneurons innerate vessels in the dorsal horn of the spinal cord of rats.

J. Physiol., 532, 749–758.

809. Zochodne, D. W., Theriault, M., Sharkey, K. A., Cheng, C., & Sutherland, G. (1997).

Peptides and neuromas: calcitonin gene-related peptide, substance P, and mast cells

in a mechanosensitive human sural neuroma. Muscle Nerve, 20, 875–880.

810. Zochodne, D. W., Thompson, R. T., Driedger, A. A. et al. (1988). Metabolic changes in

human muscle denervation: topical 31-P NMR spectroscopy studies. Magn. Reson. Med.,

7, 373–383.

811. Zuo, J., Ferguson, T. A., Hernandez, Y. J., Stetler-Stevenson, W. G., & Muir, D. (1998).

Neuronal matrix metalloproteinase-2 degrades and inactivates a neurite-inhibiting

chondroitin sulfate proteoglycan. J. Neurosci. 18, 5203–5211.

812. Zuo, J., Hernandez, Y. J., & Muir, D. (1998). Chondroitin sulfate proteoglycan with

neurite-inhibiting activity is up-regulated following peripheral nerve injury.

J. Neurobiol., 34, 41–54.

266 References

Index

A alpha sensory axon 10, 17

A beta sensory axon 10, 18, 82

A delta sensory axon 10, 18,

33, 35, 81, 142

A5E3 108

A-60 19

abaxonal Schwann cell

cytoplasm 23

acetylcholine (Ach) 37

acetylcholinesterase (AchE)

123

acetylcholine receptors (AchR)

69, 146

ACTH (adrenal corticotrophic

hormone) and

regeneration 213

actin, F-actin, G-actin 19, 97,

99, 101, 123–125, 209

actin interacting protein

87, 101

actin monomer binding

proteins 87, 101

activating transcription factor

3 (ATF3) 122–123, 193

activity, and impact on

regeneration 218

acute motor and sensory

axonal neuropathy

(AMSAN) 2

adaxonal Schwann cell

cytoplasm 23

adenomatous polyposis

coli tumor suppressor

protein (APC) 87,

102, 190

adhesive point contacts 210

adrenergic axons, terminals

13, 155–157

age, and impact on

regeneration 215

agrin 146, 176

Akt (PKB; protein kinase B;

V-akt murine thymoma

viral oncogene homolog)

87, 103–104, 120,

124, 128, 188–189,

193–194, 211

allodynia 55, 75–76, 141,

143, 150

alpha internexin 19, 21, 123

alpha motor neuron (axon) 10,

18, 25

alpha2-chimaerin 104

AMPA receptor 29

amphoterin 127

angiogenesis in peripheral

nerves 106, 166

ankyrin 19

ankyrinG 25

anterior (gray matter)

horn of the spinal

cord 6, 8, 26

arteriovenous shunts

(AV shunts) 13, 154

anterograde neuron labeling

74, 74

apolipoprotein E

Arp 2/3 complex 87, 101,

209

Artemin 191, 192–193

autocrine support of

neurons 120

autonomic nervous system

8, 31

autonomic neurons 8, 31

autoregulation 13, 27,

156–157

axon direction finding,

guidance, misdirection

115–116, 143–145, 198

axon growth rate 94

axon sprouting 71, 91, 128,

135, 140

(see also Regenerative

sprouting)

axonal endbulbs

(see Endbulbs, axonal)

axonal protein synthesis

124–126, 198

axonal pruning 93, 128–129

axolemma 19

axonotmesis 39–41, 43, 52

axoplasmic reticulum 19

267

axoplasmic transport,

anterograde, retrograde

and slow 22, 91

axotomy 6, 119, 121, 151

Bands of Bungner 92, 106, 174

basement membrane 22,

112, 207

beam walking test 78

beta-thymosin 125

blocking of neuromuscular

junction transmission

140

blood flow see nerve blood flow

blood–ganglion barrier 27

blood–nerve barrier (BNB)

14–15, 57

bone morphogenetic proteins

(BMPs) 191

brachial plexus 10

bradykinin receptor 29

brain-derived neurotrophic

factor (BDNF) 30, 98,

100, 118, 123, 127, 144,

146–148, 184–190, 191,

198–204

C unmyelinated axons 33,

81, 142

cadherins 24, 207 (see also

N-cadherin)

calcitonin gene-related

peptide (CGRP) 13–14,

29–30, 32, 63, 66, 76,

88, 90, 94, 108, 141, 143,

155, 159, 166, 173, 179,

211, 215

calcium (Ca2þ) channels 29

calcium-activated potassium

channel (KCa) 29

calmodulin 105

Ca2þ-calmodulin-dependent

protein kinase II (CaMK II)

97, 128, 209, 211

Ca2þ-calmodulin-dependent

protein kinase I

(CaMKI) 105

calpain 46, 105

campenot compartmental

cultures 80, 198

cannabanoid receptor 29

cap Z 101

cardiotropin-1 (CT-1) 190–191

casein kinase II (CK2) 103

caspase-3 124

Caspr 25, 49–50

caveolin 119

CDC 11, 34, 87, 99, 102, 155,

190, 193

celiac ganglia 8, 10, 31

cell adhesion molecules

(CAMs) 206–207

cell body reaction to

axotomy 121–122

central chromatolysis 121

central nervous system (CNS)

plasticity in response

to peripheral nerve

lesions 150

central nervous system (CNS),

regeneration of

peripheral axons into 217

CNS topographical maps 151

cervical and lumbar

enlargement 8, 10, 26

cervical ganglia 8

Charcot–Marie–Tooth disease

(CMT) 24–25

chemoreceptors 34

cholesterol 91

cholineacetyltransferase(ChAT)

11, 32, 37, 63, 123, 173

cholinergic axons,

terminals 34

chondroitin sulphate

proteoglycan (CSPG) 23,

50, 99, 102, 104, 175, 190,

200, 211–212, 217

chronic constriction injury

(CCI) 54–55, 143, 163

Chung model of neuropathic

pain 33–34, 55

ciliary neurotrophic factor

(CNTF) 118, 176–177,

190–191, 199, 202–203

cofilin 87, 101–102, 124

cold sensory testing 76, 150

collagen 11–12, 14–15, 22,

147, 174, 211

collapsin-response mediator

proteins (CRMPs) 87,

103–104

collateral sprouting of axons,

collateral reinnervation

76, 81–82, 128–130, 139,

142, 150, 172, 180, 201

compartment syndrome 39,

41, 55, 162, 166

compliance 11

compound muscle action

potential (CMAP) 43,

67–69, 138, 148,

163–164, 178

compression injuries to nerve

(see Entrapment nerve

injuries)

conduction block 40, 42–43,

164 (see also Ischemia,

conduction block)

conduction velocity 66, 68, 138

Connexin 32 (Cx32) 10, 18,

24, 25

Contactin 25

contralateral effects of

axotomy 122

Cortactin 101

cranial ganglia 8, 27

cyclic AMP (cAMP) 102, 108,

126, 146, 189, 200, 213,

217–218

cyclic AMP phosphodiesterase

126, 193

268 Index

cyclic AMP response

element-binding protein

(CREB) 109, 189

cyclin-dependent kinase

(CDK) 104

cytoskeleton-associatedprotein-

23 (CAP-23) 105, 122

demyelination 39–40, 42–43

dendritic arbours and

axotomy of motor

neurons 150–151

denervation, long-term of

nerve trunks 173–175

denervation, long-term of

nerve trunks, reasons 170

desert hedgehog protein (Dhh)

51, 87, 131

desert hedgehog protein

receptors 131

desiccation artifact of

nerve 59–60

diabetes mellitus and

regeneration 216

diabetic neuropathy 130,

162, 165, 180

diacylglycerol 189

digestion chambers, axon

or myelin 48

dishevelled proteins 62, 103

dissector, physical or

optical for unbiased

cell counts 119

dorsal horn of spinal cord 10

dorsal root entry zone

(DREZ) 217

dorsal root ganglion (DRG)

28–30, 56, 109, 157

dorsal root ganglion blood

flow 157

dorsal root ganglion, ischemia

of 165

dynein 22, 98, 126

dystroglycan 147

E-cadherin 24

(see also Cadherins)

E3/E4 ubiquitin ligases 46, 118

ectopic discharges associated

with neuropathic pain

88, 90

Edinger–Westphal nucleus 31

electrical silence of

denervated muscles 177

electrical stimulation, impact

on regeneration 144,

202, 219

electrodecremental response

in testing neuromuscular

junction 140

electromyography (EMG),

needle electromyography

69, 139–140, 178

electron microscopy (EM; thin

sections for microscopy)

18, 60, 83, 95, 106

Ena/Mena/VASP 101

endbulbs, axonal 52, 88–92

endoneurial blood flow

compartment 155

endoneurial capillaries 155

endoneurium 14–16, 155

endothelin (ET) 29, 163

endothelin receptor 29

endosomes 19

enkephalin 32, 88, 122

enteric nervous system 8, 10,

31–32, 118

entrapment, or compression

nerve injuries 56, 67, 166

ephrin 99

epidermal growth factor (EGF)

191, 194

epidermal growth factor

receptor (EGFR) 29, 99, 191

epidermis 33, 71, 140–143

(see also Skin innervation)

epineurialarterioles13,156

epineurial plexus 11, 155, 168

epineurium, epineurial

sheath 10, 12

Erb neuregulin receptors 49,

118–119, 140, 174

erythropoietin (EPO) 51,

191, 194

estrogen and regeneration 213

eukaryotic translation

initiation factor-4E

(eIF-4E) 124

evoked potentials 71

explants of neurons for

regeneration studies

80, 183

EXT2 (hereditary multiple

exostoses gene) 123

extracellular signal-related

protein kinases (ERKs)

189, 192, 194, 197

ezrin 49–50

F-spondin 145, 211

F waves 67, 70

fascicles 13, 15–16

fibrillation potentials 43, 69,

138–140, 178

fibroblast 131

fibroblast growth factors

(FGFs) 108, 118, 123, 149,

191, 193, 199–200, 203

fibroblast growth factor

receptor (FGFR) 29, 191,

193, 203

fibronectin 22, 116, 200,

209–211

filopodia of growth cones 95

fixation of nerves 61, 83

fixation artifacts of nerve

60–61

FK506-binding proteins

(FKBPs) 213

flexion contractures, paw from

prolonged denervation

79, 149, 178–179

269Index

fluorogold 72–73

focalized proteolysis 212

fodrin 19

forelimb testing of

reinnervation in rats

76, 150

forskolin and regeneration 213

G ratio 23, 62

Gp130 receptor subunit

190–191, 199

galanin 29, 66, 88, 122

galanin receptor 29

galectin-1 127

gamma motor axon 10, 18,

25, 34

ganglion blood flow 27

gangliosides and regeneration

213

gap, proximal to distal

stump following nerve

transection 64, 133–134,

201–203

gap junctions (zona adherens)

14

glial-axon junction 24

glial-derived neurotrophic

factor (GDNF) 11, 30, 118,

146, 148, 174, 186,

191–192, 197–204

glial-derived neurotrophic

factor receptors 29, 123,

191, 192–193, 199

glial fibrillary acidic protein

(GFAP) 106, 109, 113,

115, 145

glycogen synthase kinase 3b

(GSK3b) 87, 102–104, 188,

190, 209, 212

glutamate receptors 29

glycogen depletion method to

study motor units 136

glycolytic (Type II) muscle

fibers 36, 136

Glypican-1 147

GM1 ganglioside 29

Golgi organs 19, 86

Golgi tendon organs 34–35

gray rami communicans 31

grip strength testing 150

growth-associated protein 43

(GAP43/B50) 65, 87, 104,

106, 108, 122–123, 143,

173, 189, 214

growth cone 79, 88, 90, 95–96,

100, 107, 124–125

growth cone organizing

complex (GCOC) 86

growth hormone (GH) and

regeneration 213

Guillain–Barre Syndrome

(GBS) 2–3, 130

H waves, reflexes 70

hair sensory receptors 32,

35, 142, 180

hand reimplantation,

reinnervation following

141, 151

handling artifact of

histological samples 59

Hargreaves apparatus for

thermal testing 75–76

heat shock protein 27 (HSP 27)

120, 122–124

heparan (heparin) sulphate

proteoglycan 116, 123,

192–193, 211

hepatocyte growth factor

(HGF) 146, 191, 195, 199

histamine 14, 179

histamine receptor 29

HNK-1 (human natural killer

leukocyte epitope; L2)

144, 207

horizontal ladder test 79, 150

HuD (RNA binding protein)

123

hydrogen clearance

polarography (HC)

158–160

hypercarbia, impact on nerve

blood flow 156

hyperemia of nerve blood

flow 166

hyperpolarization-activated

cyclic-nucleotide gated

(HCN) 29–30

hypogastric ganglia 8, 10

IB-4 (isolectin B-4) sensory

neuron subtype 11, 30,

66, 142, 186, 192–193,

197, 211

image analysis of nerves,

ganglia 61

immunohistochemistry 11,

66, 115

immunophilins 213

importin 126

importin-vimentin-pErk

retrograde signal 126

in vitro measures of

regeneration 79

injury potentials after nerve

damage 85

inner mesaxon 23, 147

innervation ratio 27

inositol 1,4,5-triphosphate

(IP3) 189, 214

insulin 108, 118, 125, 191,

194, 203, 216

insulin-like growth factors

(IGFs) 108–109, 118, 146,

176, 191, 194, 199–200,

202–204, 216

insulin-like growth factor

receptors (IGF-R)

191, 194

insulin receptors (IRs) 191, 194

insulin receptor substrate

(IRS) 194–195

270 Index

interferon-g (IFN-g) 50

integrins 103, 109, 116,

119, 142, 147, 207,

209–211

interleukins 50, 87, 109, 118,

123, 127, 190–191,

198–199, 203

intermediate early genes 122

intermediate filament 19

internodes 16, 25, 40

intraperiod line 23, 147

iodoantipyrene [C14]

autoradiography or

distribution 158, 161

ischemia of ganglia (see dorsal

rootganglion, ischemiaof )

ischemia of nerve 11, 153,

162–165

ischemia of nerve during

injury 165, 168

ischemia, centrofascicular of

nerve 153

ischemia, nerve conduction

block 163–164

(see also resistance to

ischemic conduction

failure)

JAK (janus-kinase) and STAT

(signal transducer and

activator of transcription)

pathway 190

Jitter indicating unstable

neuromuscular

transmission 139–140

juxtaparanode 24–25

kainate receptor 29

katanin 87, 98

kinesin 22, 98

Krox20 (Egr2) 147

L1 118, 142, 207

lamellipodia 95

lamina densa 22

lamina lucida 22

laminin 22, 50, 103, 116,

147, 192, 208–211

laser doppler flowmetry (LDF)

158, 159–160, 161

leukemia inhibitory

factor (LIF) 51, 109,

118, 123, 127, 190–191,

199

LIM kinase 87, 101

LINGO 99

lipoprotein receptor-related

protein (LRP-1) 109, 195

lumbosacral plexus 10

lymphatic vessels in the

peripheral nerve trunk

12, 14, 193

lysosomes 19

macrophages 45, 48, 50, 52,

86, 127

macrophage inflammatory

protein 1a (MIP 1a) 50

macrophage stimulating

protein (MSP) 191, 195

major dense line 23, 147

mast cells 14, 16, 52, 76,

106, 131, 179, 198

matrix metalloproteinases

(MMPs) 50, 175, 212

mDia 99

mechanical sensitivity,

hyperalgesia 75–76,

141, 150

Meissner’s corpuscles (FAI)

33, 35, 142–143, 180

mesenteric ganglia 8, 10, 31

Merkel touch complexes,

domes (SAI) 32, 35,

142, 180

metabosensitive afferent

axons 77

microfilaments 19

microglia 119

microtubules 19–20, 21–86,

95–96, 98

microtubule associated

(binding) proteins

(MAPs or MBPs) 22, 98

microsphere embolization to

measure nerve blood flow

158, 161

miniature endplate potentials

(MEPPs) 37

minifascicles 15, 54, 61–62,

114–115, 130, 134,

168, 176

mitochondria 19–20

mitogen-activated protein

kinase (MAPK) 102,

118, 122, 128, 189,

193, 195, 211, 214

mixed spinal nerve 10

monocyte chemoattractant

protein-1 (MCP-1)

50, 109

mononeuritis multiplex 162

mononeuropathy 1

morphological measures

of nerve blood vessels 161

motor neurons 6, 8, 10, 25–26,

150

motor recovery 77–79, 149

motor unit 26, 70, 129, 136,

177 (see also voluntary

motor unit potentials)

motor units, giant 139–140,

177

motor unit force

measurements 70, 138

motor unit number estimation

(MUNE) 69, 138

multifiber nerve conduction

recordings 66

muscle ATPase stain for

fiber types 136

muscle atrophy 138, 177

271Index

muscle denervation 43, 69,

175–178, 179

muscle fiber type grouping

136

muscle grouped fiber

atrophy 137

muscle reinnervation 135

muscle specific kinase

(MuSK) 37

muscle spindles 34, 185

mTOR (see target (mammalian)

of rapamycin)

myelin 23–24, 24

myelin and lymphocyte

protein (MAL) 24

myelin associated

glycoprotein (MAG)

24–25, 99, 102,

200, 207

myelin basic protein (MBP)

24, 63

myelin ovoids 44

myelination, remyelination

23, 148

myelinated axon 10, 16–17,

20, 60

myelinated fiber density

18, 61

myenteric (Auerbach’s)

plexus 32

myosin in growth cones 99

myristoylated alanine-rich

protein kinase C

substrate (MARCKS) 105

N-cadherin 118, 174–175

N-WASP 101

Nancy B 2

Necdin 188

Neotrofin 204

nerve action potentials (NAPs)

66, 163

nerve blood flow 12, 153–169,

159–160, 169

nerve blood flow after

injury 166

nerve and ganglion blood flow

measurement methods

158–161

nerve growth factor (NGF)

80, 98, 100, 103, 118,

125, 128, 146–147,

182–183, 191, 197–204,

212, 214

nervi nervorum 14

nestin 145

Nests of Nageotte 119

Netrin 100, 128

neural cell adhesion molecule

(NCAM,Nr-CAM) 25,

108, 118, 144, 174–175,

192, 207

neurapraxia 39–40, 42, 171

neuregulins (NRGs) 49, 87,

108, 119, 137, 147,

175–177, 194

neurites, in vitro neuron

cultures 79, 91

neuroeffector junction 32

neurofascin 24–25

neurofilament 19–20, 20–21,

29, 36, 46, 65, 73, 86,

88, 90, 106–107, 115–117,

121–122, 208

neurogenic inflammation

76, 179

neurokinin receptor (NK1) 29

neuroma 51–52, 168, 173, 176

neuroma “in continuity”

52–54, 175

neuromuscular junction 36,

71, 104, 135, 138–139, 146

neuropathic pain 4, 54

neuropathy 1, 3–4, 12

neuropeptide Y 32, 88, 122

neuropilin receptors for

semaphorins (NP-1 and

NP-2) 104, 214

neuropoietic cytokines 118

neurotmesis 39, 41–43,

51–52, 172

neurotrophic hypothesis 119,

186–187, 196

neurotrophin-3 (NT-3) 30,

109, 123, 145–146,

184–190, 191, 198–204,

209

neurotrophin-4/5 (NT-4/5)

123, 146, 184, 191,

198–204

neurotrophin-6 (NT-6) 184

neurotrophin-7 (NT-7) 184

neurotrophins, neurotrophic

growth factors 30,

118, 182, 184–185,

191, 196–204

neurovascular bundle 11

neurturin 191, 192–193

Nexin-1 213

nicotinamide

mononucleotide

adenylyltransferase

(NMNAT) 47

Nissl substance 27, 121

nitric oxide (NO) 51, 89, 155,

166, 194

nitric oxide synthases (NOS)

51, 88, 122

nociceptin 29, 66, 88, 122

nociception 33

nociceptive 11, 18, 35, 201

nociceptors 33, 35

nodes of Ranvier 16–17, 20,

23–24, 25, 128

Nogo 24, 99, 212

Nogo receptor (NgR) 99,

128

nuclear localization signal

(NLS) 126

nuclear magnetic resonance

spectroscopy (NMRS) of

muscle 178

272 Index

nuclear transcription factor

kappa B (NFkB) 87, 102,

127, 188, 194

Oct6 (SCIP/TstI) 147

oligodendrocyte myelin

glycoprotein 99

oncomodulin 127–128

oncostatin 118, 190–191

opioid receptors 29, 88

opioid-like receptor 1, 3–4,

12, 29

outer mesaxon 23, 147

oxidative (Type I) muscle

fibers 36, 136

oxygen tensions, endoneurial

157, 161

p21-activated kinase (PAK)

87, 102

P31 NMR spectroscopy of

muscle, and

denervation 178

p38 124

p53 tumor suppressor

family 120

p75 growth factor

receptor 29–30, 98,

103, 106, 122, 142, 174,

184, 190–191, 199,

202–204

pErk 126

P0 protein (MPZ) 24, 147, 207

P2 protein 24

Pacinian corpuscles 33, 35,

142, 180

pancreatitis-associated

protein III (PAP-III) 51

para-aortic ganglia 8, 10

paracrine support of

neurons 120

paraesthesiae 141

paranodal bulbs 23

paranode 23–25

paranode–node–paranode

(PNP) 24

paraspinal sympathetic

ganglia 8, 10, 31

parasympathetic axons 31

parathyroid hormone-related

peptide (PTHrP) 109

pelvic splanchnic nerves 31

peptidergic axons, terminals

13, 155

periaxin 24

periaxonal space 23

pericytes 15

perikarya 8

perineurium 14, 131

perineuronal satellite (glial)

cell 6, 27–28, 108–109,

119, 124, 145

peripheral myelin protein 22

(PMP 22) 24

peripheral nerve trunk 4, 8, 10

peripherin 19, 21, 123–124

peroxynitrite 51

Perroncito, structures of 88

persephin 191, 192–193

PGP 9.5 (Protein gene

product 9.5) ubiquitin

hydrolase axon label 33,

65, 71, 81, 106–107, 115,

141, 143

phantom sensations 151

phosphatase and tensin

homolog deleted on

chromosome ten (PTEN)

87, 103–104, 188

phosphatidylcholine 91

phosphatidylinositol-3-kinase

(PI3K; PI3-kinase) 97,

103, 118, 120, 124, 128,

188–190, 192–195, 209,

211–212, 214

phosphatidylinositol

4,5-bis-phosphate (PIP2; PI

(2)P) 87, 101, 103, 105, 188

phosphatidylinositol

3,4,5-triphosphate (PIP3;

PI(3)P) 103–105, 188

phospholipase A2 (PLA2) 214

phospholipase C g (PLC-g)

188–189, 192, 195, 214

pilomotor axons 34

pinch test of regeneration 74

pincher pinocytotic protein

197–198

pinocytotic vesicles 14

pioneer axons 94

pituitary adenylate

cyclase-activating

polypeptide (PACAP)

29, 190

plasma extravasation 76

platelet-derived growth factor

(PDGF) 108–109, 191,

194, 204

pleiotrophin 146, 190–191,

195, 199, 204

poly ADP-ribose polymerase

(PARP) 189

polyamines and regeneration

217

polyneuronal innervation of

muscles 135

polyneuropathy 1

polyribosomes 19

polysialic acid moiety of

NCAM 144, 207

position sensibility 35

(see also proprioception)

positive sharp waves 43, 69

posterior columns of spinal

cord 10

potassium (Kþ) channels 25, 29

precapillary sphincters 13

preconditioning

(conditioning) 124, 218

preferential motor

reinnervation (PMR) 129,

144–145, 219

273Index

presynaptic Schwann cells 36

pro-NGF (pro-nerve growth

factor) 186–187, 188

profilin 87, 101

proneurotrophin 99

proprioception 4, 142

(see also position

sensibility)

protudin 91

protein kinase A (PKA) 200

protein kinase C (PKC)

104–105, 118, 120,

122, 189

pseudonerves 110

purinergic receptor 29–30, 197

RabGTPase 197

RAC 87, 99–100, 102, 104,

190, 192, 211

Ran-2 108

Rapsyn 37

Ras superfamily GTPases 99,

102, 188–189, 192–193,

195, 214 (see also RHO

family GTPases)

reactive Schwann cells 81,

106–107, 111, 113, 174

receptor for advanced

glycosylation

endproducts (RAGE)

51, 127

regeneration associated genes

(RAGs) 120, 122, 125, 172,

196, 218

regeneration conduit 64,

114–116, 133–134, 149,

200, 202

regeneration rate of

axons 170

regeneration units 95, 112

regenerative sprouting 45, 89,

92–94, 116, 128

(see also Axon sprouting)

Reich granules 22

Remak bundle 17–18,

22–23, 108

Renaut bodies 19

repetitive nerve stimulation

technique to assess

neuromuscular junctions

138–139

reperfusion injuries of

peripheral nerves 162

repulsive guidance molecule

(RGM) 215

resistance to ischemic

conduction failure

(RICF) 163

retrograde degeneration of

neurons and axons 86,

119–120, 172

(see also traumatic

degeneration)

retrograde labeling of

neurons 72–73

Rexed lamina 25

RHO family GTPases 97,

100, 102, 188, 190,

209–212

RHO kinase (ROK) 87, 99,

102, 211–212

RHOA 87, 99–100, 102, 104,

125, 211

ribosomal periaxoplasmic

plaque domains 22

ribosomal-P0 124

root avulsion 120

Ruffini endbulbs (SAII) 32, 35

S-100 immunohistochemical

marker of Schwann

cells 81

safety factor for

neuromuscular

transmission 138, 140

saltatory transmission 16

satellite cells 6, 27–28,

108–109, 119, 124, 145

Scar/WAVE 101

Schwann cell 6, 16, 22–23, 107

(see also Reactive Schwann

cells)

Schwann cell, terminal or

perisynaptic (see terminal

Schwann cells)

sciatic functional index (SFI)

77–78

Schmidt–Laterman clefts or

incisures 23, 147

Seddon classification of

nerve injuries 39, 41,

55, 162, 166

segmental demyelination 40

Semaphorin 87, 99, 102,

104, 131, 200, 211, 214

Semithin plastic (epon) light

microscopy tissue

sections 13, 26, 28, 58–59,

83, 114, 134

sensory nerve action potentials

(SNAPs) 66, 141, 149

sensory neuron 8, 10, 27–30,

96, 111–112

sensory recovery 75–77, 149

serotonin 14

Sholl analysis of neurite

outgrowth 80

single fiber electromyography

(SFEMG) 139–140

SIRT1 (Silent information

regulator mammalian

protein; sirtuin) 47

skin denervation 179

skin innervation 32–33, 71,

140–143

skin ulceration 179

(see also Wound healing,

skin)

slingshot 87, 101

sodium (Naþ) channels 25,

29, 88

somatostatin 29, 32

274 Index

somatostatin receptor 29

Sox10 (Sry-related HMG

box 10) 147

spectrin 25

sphenopalatine ganglion 31

spinal ganglia 27

spondylosis 165, 171

staggered axonal regrowth

93–95, 171

staircase test 79, 150

STAT3 (Signal Transducer

and Activator of

Transcription 3) 127

stem cells, peripheral

neuronal 145

stretch injuries 56

substance P (SP) 11, 13, 29,

62–63, 76, 88, 143, 155,

159, 179, 185

submandibular ganglion 31

submucous (Meissner’s)

plexus 32

Suc1-associated neurotrophic

factor-induced tyrosine

phosphorylated target

(SNT) 188

Sunderland classification of

nerve injuries 42

superior cervical ganglia

neural-specific 10 protein

(SCG10) 122

suppressor of cytokine

signaling (SOCS3) 190

sweat glands 34, 71, 129,

143, 179

sympathetic axons 11, 34,

155, 193

sympathetic ganglia 31

synaptophysin 106

TAJ/TROY (toxicity and JNK

inducer; member of

the TNF receptor

superfamily) 99

target (mammalian) of

rapamycin (TOR; mTOR)

124, 128

tenascin 116, 211

tensile properties 11

terminal (perisynaptic)

Schwann cells 104,

135, 142, 177, 180, 218

testosterone and

regeneration 213

thermal sensation,

hyperalgesia 18, 33, 75,

141, 143, 149–150

thyroxine (thyroid

hormone) and

regeneration 213

tight junctions (zona

occludens) 14–15

tissue plasminogen activator

(tPA) 213

transforming growth

factor-b (TGF-b) 50, 148,

191, 193, 204

transient receptor potential

proteins (TRPs) 29, 34

transitional zone (TZ) 19

traumatic degeneration 46–47

(see also retrograde

degeneration)

transient axonal glycoprotein-

1 (TAG-1; axononin-1) 24

triple response of Lewis 179

tropomyosin 101, 124

tropomyosin-related

kinase A (Trk A) 29–30,

98, 122, 147, 184–190,

197, 202

tropomyosin-related

kinase B (Trk B) 30, 98,

123, 144, 184–190, 191,

197, 202

tropomyosin-related kinase

C (Trk C) 29–30, 98,

122–123, 184–190, 191

tubulin 21, 63, 65, 80, 98,

106, 110, 115, 117,

122–123, 210

tumor necrosis factor-a

(TNF-a) 50, 109, 127

tyrosine hydroxylase (Th)

11, 32

ubiquitin 124

ubiquitin-proteasome

system (UPS) 47

uncoupling protein 3 (UCP3)

195

unmyelinated axon 10–11,

18–19

unmyelinated fiber density

19, 60

vasa nervoum 11, 15–16,

153–154, 167, 169

vascular endothelial

growth factor (VEGF)

146, 166, 175, 191,

195, 199

vasculitic neuropathy 4

vasoactive intestinal

polypeptide (VIP)

29, 32, 141, 143,

215

vasoconstriction 13–14,

155–157, 180

vasodilatation 13–14, 76,

155, 166

vesicle associated

membrane protein-1

(VAMP-1) 123

vimentin 126

voluntary motor unit

potentials 69

(see also Motor units)

voltage gated calcium

channels 36

voltage gated potassium

channels 36

275Index

von Frey hairs for

mechanical sensitivity

testing 141, 150

Wallerian degeneration and

Wallerian-like

degeneration 5, 40–43,

44–46, 48, 57, 62–63,

65, 86, 93, 118, 154,

164, 173

walking track analysis 77–78,

149–150

watershed zones 11, 153

White rami communicans 31

Wlds mouse mutant 47

wound healing, skin 179

wrong way axons 116, 143

YFP thy-1 transgenic mice

66, 93, 203

ZAS3 (Zinc finger protein,

acidic region, Ser/Thr

rich) 87, 127

276 Index

Documents

[Douglas W. Zochodne] Neurobiology of Peripheral N